Future Farming: a Return to Roots?

SC IE NTIFIC AME RIC AN AN RIC AME NTIFIC IE SC Re Future Farming

t urn andJohn P. Reganold By Jerry D. Glover, Cindy M. Cox and built deep root systems major crop plants lived for years become more sustainable if Large-scale agriculture would © 2007 SCIENTIFIC AMERICAN, INC. AMERICAN, SCIENTIFIC 2007 ©

to Ro : could could reduce those burdens. le shown on the opposite page, mediate wheatgrass and tritica the hybrid ofexperimental inter of perennial versions, such as ing Development environments. the land and polluting surround human inputs while depleting heavily on irrigation and other CROPS FOOD MODERN o t

August s? depend depend 2 07 20 - - -

CREDIT or many of us in affluent regions, our bath- Roots of the Problem room scales indicate that we get more than Most of the farmers, inventors and scientists F enough to eat, which may lead some to who have walked farm fields imagining how to believe that it is easy, perhaps too easy, for farm- overcome difficulties in cultivation probably ers to grow our food. On the contrary, modern saw agriculture through the lens of its contem- agriculture requires vast areas of land, along porary successes and failures. But in the 1970s with regular infusions of water, energy and Kansas plant geneticist Wes Jackson took a chemicals. Noting these resource demands, the 10,000-year step into the past to compare agri- 2005 United Nations–sponsored Millennium culture with the natural systems that preceded Ecosystem Assessment suggested that agricul- it. Before humans boosted the abundance of ture may be the “largest threat to biodiversity annuals through domestication and farming, and ecosystem function of any single human mixtures of perennial plants dominated nearly activity.” all the planet’s landscapes—as they still do in Today most of humanity’s food comes direct- uncultivated areas today. More than 85 percent KEY FACTS ly or indirectly (as animal feed) from cereal of North America’s native plant species, for ■ Modern agriculture’s grains, legumes and oilseed crops. These staples example, are perennials. intensive land use quashes are appealing to producers and consumers be- Jackson observed that the perennial grasses natural biodiversity and cause they are easy to transport and store, rela- and flowers of Kansas’s tall-grass prairies were ecosystems. Meanwhile tively imperishable, and fairly high in protein highly productive year after year, even as they the population will balloon and calories. As a result, such crops occupy built and maintained rich soils. They needed no to between eight billion about 80 percent of global agricultural land. But fertilizers, pesticides or herbicides to thrive and 10 billion in the com- they are all annual plants, meaning that they while fending off pests and disease. Water run- ing decades, requiring that must be grown anew from seeds every year, typ- ning off or through the prairie soils was clear, more acres be cultivated. ) ically using resource-intensive cultivation meth- and wildlife was abundant. ■ Replacing single-season this pagethis

( ods. More troubling, the environmental degra- In contrast, Jackson saw that nearby ﬁelds of crops with perennials dation caused by agriculture will likely worsen annual crops, such as maize, sorghum, wheat, would create large root sys- as the hungry human population grows to eight sunﬂowers and soybeans, required frequent and tems capable of preserving the soil and would allow JIM RICHARDSONJIM billion or 10 billion in the coming decades. expensive care to remain productive. Because ); cultivation in areas current- That is why a number of plant breeders, annuals have relatively shallow roots—most of ly considered marginal. agronomists and ecologists are working to de- which occur in the top 0.3 meter of soil—and oppositepage ( velop grain-cropping systems that will function live only until harvest, many farmed areas had ■ The challenge is monumen- much more like the natural ecosystems dis- problems with soil erosion, depletion of soil fer- tal, but if plant scientists placed by agriculture. The key to our collective tility or water contamination. Moreover, the ee- succeed, the achievement success is transforming the major grain crops rily quiet farm fields were mostly barren of would rival humanity’s original domestication of into perennials, which can live for many years. wildlife. In short, sustaining annual monocul- food crops over the past 10

NationalGeographic/Getty Images The idea, actually decades old, may take de- tures in so many places was the problem, and millennia—and be just as cades more to realize, but signiﬁcant advances the solution lay beneath Jackson’s boots: hardy revolutionary. in plant-breeding science are bringing this goal and diverse perennial root systems. —The Editors

MICHAELLEWISS. within sight at last. If annual crops are problematic and natural

www.SciAm.com SCIENTIFIC AMERICAN 83 © 2007 SCIENTIFIC AMERICAN, INC. Sept. Dec. March June Roots as Solution Today the traits of perennials are also becoming better appreciated. With their roots commonly exceeding depths of two meters, perennial plant communities are critical regulators of ecosystem functions, such as water management and carbon and nitrogen cycling. Although they do 1 m have to invest energy in maintaining enough underground tissue to survive the winter, perennial roots spring into action deep within the soil whenever temperatures are warm enough and 2 m nutrients and water are available. Their con- stant state of preparedness allows them to be highly productive yet resilient in the face of environmental stresses. In a century-long study of factors affecting soil erosion, timothy grass, a perennial hay crop, proved roughly 54 times more effective in main- PERENNIAL PLANTS, such as taining topsoil than annual crops did. Scientists intermediate wheatgrass (at right in panels above), access have also documented a fivefold reduction in nutrients and water in larger water loss and a 35-fold reduction in nitrate loss volumes of soil with their well- from soil planted with alfalfa and mixed peren- developed roots than do nial grasses as compared with soil under corn annuals, such as winter wheat and soybeans. Greater root depths and longer (at left in panels above). In turn, growing seasons also let perennials boost their perennial roots support sequestration of carbon, the main ingredient of microorganisms and other soil organic matter, by 50 percent or more as biological activity that enrich compared with annually cropped fields. Because soil. The resulting dark, they do not need to be replanted every year, pe- granular soil (far right), taken rennials require fewer passes of farm machinery from underneath a perennial meadow, retains ample water and fewer inputs of pesticides and fertilizers as and nutrients. Soil from an ecosystems offer advantages, why do none of our well, which reduces fossil-fuel use. The plants adjacent annual field (near important grain crops have perennial roots? The thus lower the amount of carbon dioxide in the right) is lighter with a weak, answer lies in the origins of farming. When our air while improving the soil’s fertility. clumped structure. Neolithic ancestors started harvesting seed-bear- Herbicide costs for annual crop production ing plants near their settlements, several factors may be four to 8.5 times the herbicide costs for probably determined why they favored annuals. perennial crop production, so fewer inputs in The earliest annuals to be domesticated, em- perennial systems mean lower cash expendi- mer wheat and wild barley, did have appealing- tures for the farmer. Wildlife also benefits: bird

ly large seeds. And to ensure a reliable harvest populations, for instance, have been shown to ) every year, the ﬁrst farmers would have replant- be seven times more dense in perennial crop

ed some of the seeds they collected. The charac- ﬁelds than in annual crop ﬁelds. Perhaps most wheatstalks teristics of wild plants can vary greatly, howev- important for a hungry world, perennials are er, so the seeds of plants with the most desirable far more capable of sustainable cultivation on

traits, such as high yield, easy threshing and re- marginal lands, which already have poor soil );GETTY IMAGES( soils sistance to shattering, would have been favored. quality or which would be quickly depleted by ( Thus, active cultivation and the unwitting ap- a few years of intensive annual cropping. plication of evolutionary selection pressure For all these reasons, plant breeders in the CornellUniversity quickly resulted in domesticated annual plants U.S. and elsewhere have initiated research and with more appealing qualities than their wild breeding programs over the past ﬁve years to de- annual relatives. Although some perennial velop wheat, sorghum, sunﬂower, intermediate );STEVE CULMAN

plants might also have had good-size seeds, they wheatgrass and other species as perennial grain roots did not need to be replanted and so would not crops. When compared with research devoted have been subjected to—or beneﬁted from—the to annual crops, perennial grain development is

same selection process. still in the toddler stage. Taking advantage of GLOVER( D. JERRY

84 SCIENTIFIC AMERICAN August 20 07 © 2007 SCIENTIFIC AMERICAN, INC. the significant advances in plant breeding over grams across the U.S. are currently pursuing the past two or three decades, however, will such interspecific (between species) and inter- make the large-scale development of high-yield generic (between genera) hybrids to develop pe- perennial grain crops feasible within the next rennial wheat, sorghum, corn, flax and oilseed 25 to 50 years. sunflower. For more than a decade, University Perennial crop developers are employing es- TOP 10 CROPS of Manitoba researchers have studied resource sentially the same two methods as those used by Annual cereal grains, food use in perennial systems, and now a number of many other agricultural scientists: direct do- legumes and oilseed plants Canadian institutions have started on the long mestication of wild plants and hybridization of claimed 80 percent of global road to developing perennial grain programs as existing annual crop plants with their wild rela- harvested cropland in 2004. well. The University of Western Australia has tives. These techniques are potentially comple- The top three grains covered already established a perennial wheat program mentary, but each presents a distinct set of chal- more than half that area. as part of that country’s Cooperative Research lenges and advantages as well. Center for Future Farm Industries. In addition, CROP LAND % scientists at the Food Crops Research Institute Assisted Evolution 1. Wheat 17.8 in Kunming, China, are continuing work initi- Direct domestication of wild perennials is the 2. Rice 12.5 ated by the International Rice Research Insti- more straightforward approach to creating tute in the 1990s to develop perennial upland 3. Maize 12.2 perennial crops. Relying on time-tested meth- rice hybrids. 4. Soybeans 7.6 ods of observation and selection of superior indi- At the Land Institute, breeders are working vidual plants, breeders seek to increase the fre- 5. Barley 4.7 both on domesticating perennial wheatgrass quency of genes for desirable traits, such as easy 6. Sorghum 3.5 and on crossing assorted perennial wheatgrass separation of seed from husk, a nonshattering 7. Cottonseed 2.9 species (in particular, Th. intermedium, Th. seed, large seed size, synchronous maturity, pal- 8. Dry beans 2.9 ponticum and Th. elongatum) with annual atability, strong stems, and high seed yield. wheats. At present, 1,500 such hybrids and 9. Millet 2.8 Many existing crops, such as corn and sunflow- thousands of their progeny are being screened 10. Rapeseed/mustard 2.2 ers, lent themselves readily to domestication in for perennial traits. The process of creating this manner. Native Americans, for example, these hybrids is itself labor-intensive and time- turned wild sunflowers with small heads and consuming. Once breeders identify candidates seeds into the familiar large-headed and large- for hybridization, they must manage gene ex- seeded sunflower [see box on page 88]. changes between disparate species by manipu- Active perennial grain domestication pro- lating pollen to make a large number of crosses grams are currently focused on intermediate between plants, selecting the progeny with de- wheatgrass (Thinopyrum intermedium), Max- sirable traits, and repeating this cycle of cross- imilian sunflower (Helianthus maximiliani), Il- ing and selection again and again. linois bundleflower (Desmanthus illinoensis) Hybridization nonetheless is a potentially and flax (a perennial species of the Linum ge- faster means to create a perennial crop plant nus). Of these, the domestication of intermedi- than domestication, although more technology ate wheatgrass, a perennial relative of wheat, is is often required to overcome genetic incompat- perhaps in the most advanced stages. ibilities between the parent plants. A seed pro- To use an existing annual crop plant in creat- duced by crossing two distantly related species, ing a perennial, wide hybridization—a forced [THE AUTHORS] for example, will often abort before it is fully mating of two different plant species—can bring developed. Such a specimen can be “rescued” as Jerry D. Glover is an agroecolo- together the best qualities of the domesticated an embryo by growing it on artificial medium gist and director of graduate annual and its wild perennial relative. Domes- research at the Land Institute in until it produces a few roots and leaves, then ticated crops already possess desirable attri- Salina, Kan., a nonprofit organiza- transferring the seedling to soil, where it can butes, such as high yield, whereas their wild rel- tion devoted to education and grow like any other plant. When it reaches the atives can contribute genetic variations for traits research in sustainable agriculture. reproductive stage, however, the hybrid’s genet- Cindy M. Cox is a plant patholo- such as the perennial habit itself as well as resis- ic anomalies frequently manifest as an inability gist and geneticist in the insti- tance to pests and disease. tute’s plant-breeding program. to produce seed. Of the 13 most widely grown grain and oil- John P. Reganold, who is Regents A partially or fully sterile hybrid generally re-

);SOURCE: FAOSTAT seed crops, 10 are capable of hybridization with Professor of Soil Science at Wash- sults from incompatible parental chromosomes perennial relatives, according to plant breeder ington State University at Pullman, within its cells. To produce eggs or pollen, the soybeans specializes in sustainable agricul- T. Stan Cox of the Land Institute, a Kansas non- hybrid’s chromosomes must line up during mei- ture and last wrote for Scientiﬁc proﬁt that Jackson co-founded to pursue sus- American on that subject in the osis (the process by which sex cells halve their

GETTYIMAGES( tainable agriculture. A handful of breeding pro- June 1990 issue. chromosomes in preparation for joining with

www.SciAm.com SCIENTIFIC AMERICAN 85 © 2007 SCIENTIFIC AMERICAN, INC. [BENEFITS] SUSTAINABLE FARMING: NEW VS. NOW The potential advantages of future perennial crop plants are visible today by comparing perennial wheatgrass (below left) growing alongside domesticated annual wheat (below right). Although a perennial wheat could one day yield grains similar to those of the annual crop, it might live for many years and look much more like its wheatgrass relative belowground. Perennial crops would transform the process of farming and its environmental effects by using resources more effectively, thereby being less dependent on human inputs and more productive for a longer time. Perennials also anchor and support the ecosystem that nourishes them, whereas short-lived and short-rooted annuals allow water, soil and nutrients to be lost. Experimental perennial wheat Photosynthesis takes up atmospheric carbon PERENNIAL

After seed harvest, livestock could graze vegetation

Competitive roots discourage weeds with less use of herbicide

Wildlife thrives in plant shelter CARBON FACTOR Seasonal regrowth Global warming potential— from roots or rhizomes lengthens productive period greenhouse gases released into the atmosphere by crop production inputs, minus carbon sequestered in soil—is negative for perennial crops. The more resilient perennials are also expected to fare better than annuals in a warming climate.

SOIL CARBON SEQUESTERED (kilograms per hectare per year) Annual crops 0 to 450 Perennial crops 320 to 1,100 Roots capture and use more rainwater GLOBAL WARMING POTENTIAL (kilograms of CO2 equivalent per hectare per year) Annual crops 140 to 1,140 Perennial crops –1,050 to –200 Diverse perennial crop types could share a ﬁeld, with their roots tapping different soil levels ESTIMATED IMPACT ON YIELD OF 3º C TO 8º C TEMPERATURE INCREASE (megagrams per hectare) Annual crops –1.5 to –0.5 Perennial crops +5

Roots descending two meters or more leak carbon- rich plant sugars into soil, feeding organisms that create and manage other nutrients. Additional carbon is sequestered within the roots

86 SCIENTIFIC AMERICAN August 20 07 © 2007 SCIENTIFIC AMERICAN, INC. another gamete) and exchange genetic informa- SUSTAINABLE FARMING: NEW VS. NOW tion with one another. If the chromosomes can- not find counterparts because each parent’s ver- sion is too different, or if they differ in number, Multiple passes of machinery in spring and the meiosis line dance is disrupted. This prob- fall to plow seedbeds, fertilize soil, plant seeds and apply herbicides use fossil lem can be overcome in a few ways. Because fuels and generate carbon dioxide sterile hybrids are usually unable to produce male gametes but are partially fertile with fe- male gametes, pollinating them with one of the original parents, known as backcross- ANNUAL ing, can restore fertility. Doubling the number of chromosomes, either spontaneously or by adding chemicals such as colchicine, is another strategy. Although each method allows for chromosome pairing, subsequent chromosome eliminations in each successive generation often happen in perennial wheat hybrids, particularly to chromosomes inherited from the perennial parent. Because of the challenging gene pools creat- ed by wide hybridization, when fertile perennial hybrids are identified, biotechnology techniques that can reveal which parent contributed Small roots provide less access to water parts of the progeny’s genome are useful. One and nutrients and sequester little carbon of these, genomic in situ hybridization, for example, distinguishes the perennial parent’s chromosomes from those of the annual parent by color fluorescence and also detects chromosome anomalies, such as structural rearrange- ments between unrelated chromosomes [see Topsoil and applied chemicals run off into waterways, increasing silt bottom illustration on next page]. Such analyt- and polluting drinking water ical tools can help speed up a breeding program once breeders discover desirable and undesir- able chromosome combinations, without com- promising the potential for using perennial grains in organic agriculture, where genetically engineered crops are not allowed. Another valuable method for speeding and Soil nutrients are lost along improving traditional plant breeding is known with up to 45 percent of as marker-assisted selection. DNA sequenc- annual rainwater es associated with specific traits serve as Nitrogen released into waterways markers that allow breeders to screen promotes marine dead zones crosses as seedlings for desired attributes without having to wait until the plants grow to maturity [see “Back to the Future of Cereals,” by Stephen A. Goff and John M. Salmeron; Scientific American, August 2004]. At present, no markers specific to perennial plant breeding have been established, al- Short growing season gives plants little time to capture sunlight or though it is only a matter of time. Scientists at participate in ecosystem. Fields Washington State University, for example, have can remain barren much of the year already determined that chromosome 4E in Th. elongatum wheatgrass is necessary for the important perennial trait of regrowth following a sexual reproduction cycle. Narrowing down JIM RICHARDSON (soil cross section); THE LAND INSTITUTE (insets on opposite page); JIM RICHARDSON (farm machinery); SEAWIFS PROJECT (NASA/GSFC) AND GEOEYE (dead zone); KEN CEDENO (tilled land); JACK DYKINGA USDA/ARS (erosion) SCIENTIFIC AMERICAN 87 © 2007 SCIENTIFIC AMERICAN, INC. [THE NEXT STEP] CREATING A NEW CROP To develop high-yield perennial crop plants, scientists and breeders can domesticating the small-seeded wild annual sunflower (a) into the either domesticate a wild perennial plant to improve its traits or hybridize modern annual crop plant (b) by selecting and cultivating plants with an annual crop plant with a wild perennial relative to blend their best desirable traits, such as large seeds and yields. Efforts are currently under qualities. Each method requires time- and labor-intensive plant way to directly domesticate wild perennial sunflower species (c) and also crossbreeding and analysis. Native Americans spent thousands of years to produce hybrids of the modern annual and wild perennials (d).

a b c d

Experimental hybrid of annual Wild annual Domesticated Wild perennial sunflower and a wild sunflower annual sunflower sunflower perennial relative

the region on 4E to the gene or genes that pro- high seed yield often focus on such physiologi- duce the trait would reveal relevant DNA mark- cal trade-offs, assuming that the amount of car- ers that will save breeders a year of growing bon available to a plant is fixed and therefore time in assessing hybrids. that carbon allocated to seeds always comes at Perennialism is nonetheless an intricate life the expense of perennating structures, such as path that goes well beyond a single trait, let alone roots and rhizomes. Doubters also often over- a single gene. Because of this complexity, trans- look the fact that the life spans of perennial genic modification (insertion of foreign DNA) is plants exist along a spectrum. Some perennial unlikely to be useful in developing perennial prairie plants may persist for 50 to 100 years, grains, at least initially. Down the road, trans- whereas others live for only a few years. Fortu- genic technology may have a role in refining sim- nately for breeders, plants are relatively flexible ple inherited traits. For example, if a domesticat- organisms: responsive to selection pressures, ed perennial wheatgrass is successfully devel- they are able to change the size of their total car- oped but still lacks the right combination of bon “pies” depending on environmental condi- gluten-protein genes necessary for making good- tions and to change the allocation of pie slices. quality bread, gluten genes from annual wheat A hypothetical wild perennial species might could be inserted into the perennial plant. live 20 years in its highly competitive natural environment and produce only small amounts Trade-offs and Payoffs of seed in any year. Its carbon pie is small, with Although perennial crops, such as alfalfa and much of it going toward fending off pests and

sugarcane, already exist around the world, none disease, competing for a few resources and per- ) CHROMOSOMES of an experimen- has seed yields comparable to those of annual sisting in variable conditions. When breeders tal hybrid perennial wheat plant grain crops. At first glance, the idea that plants take the wild specimen out of its resource- chromosomes are tagged with fluorescence can simultaneously direct resources to building strapped natural setting and place it into a man- to reveal whether they originat- and maintaining perennial root systems and aged environment, its total carbon pie suddenly ed with the hybrid’s wheat- also produce ample yields of edible grains may grows, resulting in a bigger plant. ( COX M.); CINDY grass (green) or wheat (red) parent. This technique helps to seem counterintuitive. Carbon, which is cap- Over time, breeders can also change the size sunflowers identify desirable chromosome tured through photosynthesis, is the plant’s of the carbon slices within that larger pie. Mod- combinations and highlights main building block and must be allocated ern Green Revolution grain breeding, when anomalies, such as fused among its various parts. combined with increased use of fertilizers, more

chromosomes (arrows). Critics of the idea that perennials could have than doubled the yield of many annual grain LANDTHE INSTITUTE (

88 SCIENTIFIC AMERICAN August 20 07 © 2007 SCIENTIFIC AMERICAN, INC. crops, and those increases were achieved in er biodiversity would in turn benefit the envi- BREEDING HYBRID plants can plants that did not have perennating structures ronment and the farmer’s bottom line. require rescuing an embryo to sacrifice. Breeders attained a portion of those Global conditions—agricultural, ecological, from the ovary (left). A impressive yield expansions in annual crops by economic and political—are changing rapidly researcher bags annual sor- selecting for plants that produced less stem and in ways that could promote efforts to create pe- ghum heads to collect pollen, with tall perennial sorghum in leaf mass, thereby reallocating that carbon to rennial crops. For instance, as pressure mounts the background (right). seed production. on the U.S. and Europe to cut or eliminate farm Yields can be similarly increased without subsidies, which primarily support annual crop- eliminating the organs and structures required ping systems, more funds could be made avail- for overwintering in perennial grain crops. In able for perennials research. And as energy pric- ➥ MORE TO fact, many perennials, which are larger overall es soar and the costs of environmental degrada- EXPLORE than annuals, offer more potential for breeders tion are increasingly appreciated, budgeting Perennial Grain Crops: An Agri- to reallocate vegetative growth to seed produc- public money for long-term projects that will re- cultural Revolution. Edited by tion. Furthermore, for a perennial grain crop to duce resource consumption and land depletion Jerry D. Glover and William Wilhelm. be successful in meeting human needs, it might will become more politically popular. Special issue of Renewable Agricul- need to live for only five or 10 years. Because the long timeline for release of pe- ture and Food Systems, Vol. 20, No. 1; March 2005. In other words, the wild perennial is unnec- rennial grain crops discourages private-sector essarily “overbuilt” for a managed agricultural investment at this point, large-scale government Wes Jackson (35 Who Made setting. Much of the carbon allocated to the or philanthropic funding is needed to build up a Difference). Craig Canine in plant’s survival mechanisms, such as those al- a critical mass of scientists and research pro- special anniversary issue of Smithson- lowing it to survive infrequent droughts, could grams. Although commercial companies may ian, Vol. 36, No. 8, pages 81–82; November 2005. be reallocated to seed production. not profit as much by selling fertilizers and pes- ) ticides to farmers producing perennial grains, Prospects for Developing Peren- belowright Greener Farms they, too, will most likely adapt to these new nial Grain Crops. Thomas S. Cox, Thus, we can begin to imagine a day 50 years crops with new products and services. Jerry D. Glover, David L. Van Tassel, from now when farmers around the world are Annual grain production will undoubtedly Cindy M. Cox and Lee D. DeHaan in BioScience, Vol. 56, No. 8, pages 649– );GETTY IMAGES( walking through their fields of perennial grain still be important 50 years from now—some 659; August 2006. right crops. These plots would function much like the crops, such as soybeans, will probably be diffi- above Kansas prairies walked by Wes Jackson, while cult to perennialize, and perennials will not Sustainable Development of the also producing food. Belowground, different completely eliminate problems such as disease, Agricultural Bio-Economy. Nicho- types of perennial roots—some resembling the weeds and soil fertility losses. Deep roots, how- las Jordan et al. in Science, Vol. 316, pages 1570–1571; June 15, 2007.

); JIM RICHARDSON);JIM ( long taproots of alfalfa and others more like the ever, mean resilience. Establishing the roots of eft thick, fibrous tangle of wheatgrass roots— agriculture based on perennial crops now will The Land Institute: abovel would coexist, making use of different soil lay- give future farmers more choices in what they www.landinstitute.org ers. Crops with alternative seasonal growth can grow and where, while sustainably produc- habits could be cultivated together to extend the ing food for the burgeoning world population g THE LANDTHE INSTITUTE ( overall growing season. Fewer inputs and great- that is depending on them.