Investing in Perennial Crops to Sustainably Feed the World

The dramatic increases in yields of annual crops are approaching their limits. But similar advances are possible in hundreds of underused perennial species.

The world’s food supply is insecure and inadequate and growing more so. But that gloomy prospect could be altered dramatically if the world adopted a novel but simple strategy: supplement the annual food crops that will soon be unequal to the task of sustaining us with improved perennial plants such as food-bearing and bioenergy-producing trees, shrubs, forbs, and grasses. Making this shift will not be easy and will require significant additional research. But we believe it is a practical and relatively inexpensive approach that will not only increase food and energy security, but will also improve soil quality, protect water resources, reduce floods, harvest carbon dioxide (CO2), and provide jobs for millions of people.

During the past half-century, the world’s population has doubled, increasing by more than three billion people. Fortunately, grain production has more than doubled during that time. Availability of nutritious food at the lowest cost in history underlies whatever peace, prosperity, and progress human society has enjoyed.

In the next half-century, however, the planet is scheduled to add another three billion people. We will once again need to increase food production by an equivalent amount just to stay in place. But staying even will not be enough. Today there are millions of people who don’t have enough to eat and millions more who yearn to move away from mainly plant-based diets. The recent push to produce crops for energy use complicates the situation.

Increases in the yields of the annual crops on which we now rely are certainly possible and should be pursued vigorously. However, these efforts are unlikely to provide a complete and sustainable solution to the growing scarcity of food and energy. Much of the land on which we depend is losing productivity because of deforestation, development, overgrazing, and poor agricultural practices. Erosion, pollution, and the expansion of deserts are among the consequences.

Water tables are falling as aquifers are pumped at rates exceeding their ability to recharge. Even the water in deep-fossil aquifers, laid down millions of years ago and which can’t be recharged, is being depleted. Nearly 90% of all fresh water used by humans goes for irrigation. According to the United Nations Food and Agriculture Organization (FAO), just 16% of the world’s cropland is irrigated, but this 16% produces 36% of the global harvest.

The stripping of forest and grassland and the cultivation of sloping land have led to rapid runoff of rainwater that normally would help recharge near-surface aquifers. In many regions, inadequate drainage has increased the salt content of the soil, leading to a loss of productivity and sometimes abandonment of agriculture altogether. The once-fertile crescent of the Middle East is a striking example, and similar salinization is accelerating in the United States, China, and elsewhere. It is certainly possible and imperative to increase the efficiency of agricultural water use, but it is not clear whether this will fully compensate for water losses or increase yields of annual crops enough.

Dust bowls and desertification are serious in many parts of the world. Depletion of the fossil aquifer under the North China plain, for example, has led to huge dust storms that choke South Koreans every year. Increasingly frequent storms from Africa routinely drop irreplaceable soil into the Caribbean, endangering the coral and thus the ecosystem there while depleting African lands.

Worldwide cereal crop production appears to have leveled off during the past few years, and per capita cereal crop production has been declining since the 1980s. One result of this decline has been steadily increasing pressure to convert environmentally sensitive land to annual crop production. A vicious cycle of burgeoning population and diminishing soil productivity has led to farming marginal land.

In addition, tropical forests are being cut for timber and burned to create grassland for cattle and farmland for soybeans and other annual crops. But many tropical and subtropical soils are fragile and cannot sustain nontree crops even with regular supplies of increasingly expensive nutrients from petroleum-based fertilizers.

When rainforests are depleted, the climate is altered because moisture, no longer retained in the standing biomass and soil, runs back to the ocean. Whereas evaporation and transpiration from tall, deep-rooted trees normally lead to further rainfall inland, depleted forests become more and more dry, reducing productivity and increasing fire danger.

The problems above are compounded by the increasing diversion of land and water to nonagricultural uses such as factories, residences, and other development. Countries that had been self-sufficient in grain production or even exported it now face declining harvests and need to import food. It takes roughly 1,000 tons of water to grow a ton of grain, so a country that imports another’s grain is also importing its water. Saudi Arabia, China, and other countries are buying or leasing large tracts of land in South America, Africa, Australia, and elsewhere to grow food, often displacing local people who depend on that land.

Social unrest and increased political conflict over shortages of food, energy, and water will be among the likely results. There is evidence that a significant contributing factor to the genocide in Rwanda was the pressure put on the land by rising population and diminished productivity, leading to social decay and murderous political instability. Food riots in a number of countries are now reported frequently. Rising food prices have been cited as contributing to the revolutions in Egypt and Tunisia and to unrest in other countries, and the price increases appear to be driven by long-term trends rather than caused by one or a few unusual events.

The perennial solution

We propose a conceptually simple approach that could make a serious dent in these problems: plant more perennial crops, whose genetic potential and agronomic possibilities have barely been tapped. There are hundreds of species of perennial grasses, shrubs, grains, legumes, and trees that could be selected, improved, and grown on hundreds of millions of hectares of damaged land. The result would be a dramatic increase in high-protein food for people and livestock and also wood and other biomass for fuel and construction. Many perennial plant species could be converted to biodiesel, ethanol, biochar for soil enrichment, and other useful products, even plastics.

The benefits would not stop with more food and biomass. Perennial crops would increase soil organic matter, reduce pollution, and stabilize soils against erosion. They would help fields, forests, and rangelands retain water, thereby reducing flooding and helping aquifers recharge. Perennials would also sequester large quantities of CO2, helping to slow climate change.

Our approach is essentially an adaptation of the Green Revolution. Its advances, which have largely held food insecurity at bay in recent decades, were based on hybrid seeds of high-yielding annual cereal grains, plus pesticides, synthetic fertilizers, and irrigation. The ecological price has been high, however, and current agricultural practices are not sustainable.

Because of their high concentrations of nutritious protein and healthy fats, tree nuts are excellent substitutes for or complements to meat in the human diet.

To be sure, the further development of existing annual crops is not only possible but imperative, especially in Africa. R&D should be intensified. Still, the Green Revolution’s dramatic increases in crop yields and feed efficiency may be approaching their limits. Fortunately, similar advances in productivity could be made in hundreds of underused perennial species, as already demonstrated in oil palm, rubber, and eucalyptus.

Many Green Revolution techniques could be applied to perennial crops, such as classical plant breeding to improve yield, stress tolerance, disease resistance, and other characteristics. But Green Revolution plants typically achieved their high yields via fertilizer, irrigation water, and pesticides. The aim of our proposal would be to develop perennial crop plants that require as little external input as possible.

If perennial plants have so much potential, why have they not received more attention? One reason is that the Green Revolution’s massive improvements in the food supply made it seem best to invest in further development of annuals. Another is that it takes relatively few years to develop improved strains of annual plants. New varieties of perennials take longer. It is often necessary to observe the results of crossbreeding for several years to know whether a new plant strain is really an improvement. Technology can speed up that process, fortunately. The newer laboratory methods, including high-throughput sequencing and comparative genomics, can more rapidly determine whether a given cross has the desired characteristics.

What to plant

Tree plantings would be of three types: nut and other food-bearing trees, oil palm, coconut, and other perennial oil-producing plants for fuel and food, and fast-growing species such as poplar and eucalyptus for rapid production of woody biomass. Nut-producing trees include species adapted to many different climates. Because of their high concentration of nutritious protein and healthy fats, tree nuts are excellent substitutes for or complements to meat in the human diet. (At current yields, one hectare of walnuts alone could supply 10% of a 2,000-calorie daily diet for 47,000 people.) Nuts also offer opportunities for improvement via classical plant breeding, complemented as appropriate by biotechnology. Research should focus on improvements in yield, nutrition, adaptation to different soils and environments, pest resistance, stress tolerance, and timber qualities. It is important to realize that food security comes not from one single source of nourishment but from a multiplicity, especially in the face of rising prices and growing climatic instability.

Some substitution of nuts for meat could have significant environmental benefits. Current world production of more than 75 million metric tons of meat from beef, sheep, pigs, and goats using current livestock management techniques often results in overgrazing, causing rangeland deterioration, especially in developing countries. This deterioration has been associated with severe erosion, soil compaction, violent sandstorms, desertification, and reduced harvest. Rapid population growth and the resulting increase in meat consumption will accelerate these problems, although improved varieties of livestock and grasses and better management practices, including rational and rotational grazing, would greatly increase grassland productivity and soil quality.

More than two billion of the world’s people acquire their energy for heating and cooking from the burning of wood, crop residues, and animal manure, and that is unlikely to change any time soon. In addition to nut trees, therefore, fast-growing nonfood trees would be needed. These could be intercropped among food-bearing plants, promoting a genetically diverse perennial polyculture. More fuel trees also would mean that crop residues and manure that are now burned for fuel could instead be used for enriching the soil.

Wide implementation of constantly improving strains of oil palm and improved agronomy could help supply fuel as well as food.

Perennial grasses, shrubs, and forbs should be planted as polycultures on land unable to sustain trees and would also be used as ground cover and among tree plantings. Because some of these plants are legumes, they would enrich the soil with nitrogen. Perennial legumes and grasses are also being developed to yield edible grains that could be harvested every year without replanting.

Perennial grasses and in many cases trees can be planted on mountainsides and other sloping land that is not suitable for annual crop cultivation but has been cultivated and degraded nonetheless. There is, in fact, more area suitable for production of perennial food, pasture, and bioenergy-producing crops available on steep and marginal land than on the land currently being used for sustainable production of cultivated annual crop plants. Grasses, legumes, and hardy native “weeds” could in many cases serve to restore degraded soil to the point where other perennial plants, including trees, could grow.

Perennials can also help reduce CO2. Fuel from biomass is often said to be carbon neutral because CO2 converted into biomass is subsequently returned to the atmosphere. But it is really better than carbon neutral because some of the carbon fixed by plants, along with other nutrients, is added to the soil, thereby enriching the productivity of the land.

By planting improved perennials, it would be possible to produce many times the biomass originally harvested from plants on degraded lands and thus reduce atmospheric CO2. This would be achieved by harvesting existing perennial plants at intervals of several years and replacing them with much more productive varieties developed by plant breeders. Most important, harvesting plants just at the peak of their growth would add greatly to biomass production because mature plants generally show little net growth and harvest of CO2. Planned harvesting and replanting, including soil enrichment by cycling some of the harvest back into the soil, would thus maximize yield and the fixation of atmospheric carbon. These ideas could be implemented with currently available technologies and are sustainable into the distant future.

Currents efforts

There are successes in the development of perennial crops that we can build on. In a set of reports entitled “Trees Outside Forests,” the FAO has described, for example, how farmers in the Sahel in West Africa leave trees scattered in their fields of annual crops in what is called parkland agriculture. The trees provide food during the dry season and longer-term droughts, as well as other products. The sheanut tree (also known as karaté) is prized for producing oil used in cooking. The oil is also used in expensive face creams sold on international markets. Fruit and seeds from baobab trees provide vitamin C, and the leaves are used in sauces.

But the sheanut and other trees grown by the Sahel farmers are hard to propagate. Few seeds germinate, and there is much potential scope for improvement in yield, drought tolerance, and other characteristics. Because these trees are already valued by the local people, integrating improvements should, with reasonable preparation, go smoothly.

Wherever successful development has taken place, researchers have worked closely with the local farmers and farmers’ associations, local governmental and university staff, and with NGOs. We think the same model would apply to developing perennial crops. Parkland improvement is only one area to be explored. Forest and orchard work, windbreaks, village-owned plantations, and many other forms of development have received attention and could use more.

Change along the lines suggested here is taking place in Europe, the United States, New Zealand, Iceland, South Korea, and China among other places. For example, large-scale plantations of perennial sea buckthorn shrubs on more than 1.2 million hectares in northwest China have reduced soil erosion and land degradation and have formed a new sustainable sea buckthorn industry for the local economy. Buckthorn berries are high in vitamin C and are used to make nutritious juice and jam, as well as face creams and medicinal products. The plant also fixes nitrogen.

Wide implementation of constantly improving strains of oil palm and improved agronomy could help supply fuel as well as food. No additional tropical forests would need to be taken for the production of added food and biodiesel from palm oil or coconuts. Intercropping oil palm trees with other perennial crops to avoid large-scale monocultures would increase the land area needed to produce a given amount of oil, but there is more than enough degraded land available. In addition, polyculture would mimic some of the diversity of healthy ecosystems and could provide improved habitat for wildlife. What is more, palm-oil production yields more than 20 tons of additional dry biomass per hectare per year, some of which could be used as an additional bioenergy source and some to enrich the soil.

The benefits would be more than ecological. As fossil fuels become more expensive, food costs will rise because current agricultural practices require large inputs of fossil fuel energy. In fact, food costs are already rising because of competition between the use of agricultural output for food and energy.

The approach described here would provide employment for large numbers of people, reducing poverty in the developing and developed worlds. It would also reduce dependence on petroleum. In addition, biofuels produced in developing countries would reduce fuel imports, stimulate the countries’ economies, and reduce the price of crude oil worldwide.

Bioethanol and biodiesel from annual crops such as corn, wheat, and oil rape often use nearly as much energy as they save because of the large amounts of fossil fuels required for production, transportation, and processing. In addition, variable amounts of soil carbon are lost to oxidation and erosion, and nitrogen oxides (potent greenhouse gases) are emitted.

In contrast, biofuels and other bioproducts produced from perennial crops such as oil palm, jatropha, switchgrass, miscanthus, sugar cane, poplar, hybrid willow, and eucalyptus use much less energy in their production, add organic carbon to the soil, reduce or eliminate soil erosion by maintaining ground cover, and have a longer period of photosynthesis during the year, thereby fixing more carbon over time.

Oil palm cultivation for biodiesel has recently been criticized because of destructive practices such as monocrop plantings, cultivation on unsuitable land leading to ecological deterioration, rapacious cutting or burning of virgin forests, and other practices that release significant amounts of heat-trapping gases. Like any agricultural activity, oil palm cultivation can be done well or badly, but it need not be done badly.

How to start

The first step toward increasing the proportion of perennial crop plants is to establish several agricultural research stations to study varieties of trees, grasses, and other perennial plants adapted to the local climate, soil, and people’s cultural practices. Studying local soil types and finding ways of improving them are critical.

Three to five stations each would be needed in China, South Asia, Africa, South America, Europe, Japan, Australia, North America, and the Middle East, or 27 to 45 stations worldwide. An initial endowment of $20 million to $40 million for each station would yield investment income of between $1 million ($20 million at 5%) to $2.8 million ($40 million at 7%) per year. This would be enough to initiate and run a station up to the point at which implementation of the plan it develops could become a reality.

The total cost of between $540 million and $1.8 billion would be incredibly inexpensive by world budgetary standards. All of it could be supplied over several years by non-governmental sources such as foundations and individuals.

We aim for private self-sustaining support because government support is often unreliable and cannot be counted on for long term R&D, and private industry is focused on short-term return. Support from governments in many countries would eventually be necessary as large-scale implementation begins, but initial support from foundations and individuals would give the research stations time to develop political networks that could eventually garner government support.

Many of the stations could be additions to or outgrowths of existing research stations and educational institutions, where infrastructure is already in place. Some could be newly created freestanding entities. One healthy byproduct of that approach would be to build R&D capacity in countries that need it.

In addition to creating useful plant varieties through long-term breeding programs, stations would initiate the agronomic, ecological, cultural, and economic analyses needed to implement sustainable local programs. They would work closely with local populations, along the lines of U.S. agricultural extension services. Local people should be brought into the planning process early, because without them any plans will probably founder. The effectiveness of local farmers’ participation in Africa and elsewhere has been well documented.

We suspect that nut and other tree cultivation would yield ecological, dietary, and economic benefits, as in the case for walnuts, but the staff of the stations envisaged here would have to examine questions such as these and many others in detail. The oil palm example illustrates some of the factors that would have to be considered for any crop in any location.

In addition, many dozens to hundreds of species of tree, grass, and forb would have to be evaluated; we have mentioned only a few of the possibilities. No one station could work on hundreds of species. A particular station might work on perhaps a dozen, depending on local soil conditions, topography, and other factors. That is one reason why many stations would be necessary. These species would have to be studied in groups to avoid monocultures, groups whose compositions would vary from place to place. This would enhance biodiversity and maximize economic and environmental benefits. By comparison, annual crop agriculture depends primarily on a small number of species. Worldwide coordination to facilitate the exchange of germplasm as well as information would be mostly informal among the stations, with some help from international organizations such as the FAO, the World Food Program, the Consultative Group on International Agricultural Research (CGIAR), universities, and national governments. No new worldwide bureaucracy need be created.

A major area of inquiry for the stations would be ecological questions such as what the yield might be on hillsides and other types of land and would vary with locale. It might be desirable to accept lower yields on hillsides, for example, if that would help restore the soil or reduce flooding and landslides. Yields lower than those on the best bottomland soils would be better than no yield at all, and yields would rise as soils improve and the land is replanted with superior varieties developed at the stations. Throughout their work, the stations would employ the principles of restoration ecology.

We are aware of the potential ecological damage that could arise from ill-considered plantings of potentially invasive species or contaminated seed. However, nearly all of the annual crops on which we currently depend for food are exotics in most of the places they are grown. This is also true of agriculturally valuable perennials in current use worldwide. The fact that a species is nonnative does not by itself disqualify it as a potential component of a sustainable perennial cropping system. Potential ecological risks and other environmental problems would be a major area of investigation.

As we have said, the amount of money these projects would need is not large by world standards, and the funds can be raised over several years. When stations are set up, especially in poor countries, the principal dangers will be corruption and political interference. Some countries may have to be avoided, at least at first. Networking to find people who are knowledgeable about local conditions will have to take place. Ties to local universities and agricultural agencies will have to be forged to encourage joint work by faculty, staff, and students. The stations could also help train people from other countries through formal programs, internships, and the like. Existing organizations, such as the CGIAR, the World Agroforestry Centre, and others can assist. Close and continuing contact would have to be established with them, with other NGOs, and with ministries of national governments. Fortunately, they are open to such contact. It would be useful, especially in the beginning, to augment the work of these and other centers directly by adding staff and programs to develop perennial plants as mentioned above. As funds are raised and specific regional needs are identified, additional stations could be set up or existing ones expanded.

There are many possible models of how to set up these research stations and how they might operate. Here is one potential scenario: With money available, a station would need a few people trained in plant breeding and in agronomy. These could be either local people or expatriates recruited through networking. Local farmers could also be hired to work. They would provide not just labor but empirical expertise and cultural knowledge on which the scientific staff could draw. A person trained in laboratory work would be essential to speed determination of whether a given cross of two plant strains has the characteristics desired. Work would begin with efforts to improve a few species already in local use while simultaneously testing and adapting a few from outside the local area. The latter would be chosen because the local soil, topography, and climate appear suitable and because they are likely to fit within local dietary and economic needs. Soil and water surveys would be undertaken from the start. One or two people trained in agricultural economics and others who understand the area’s cultures and politics would also be hired. The working staff of a functioning station might thus involve two plant breeders/agronomists, a laboratory worker unless one of the breeders has those skills, perhaps half a dozen farmers full and/or part time, an economist, and a cultural/political analyst. Some of the staff could be shared with other local organizations. This comes to perhaps 10 full- or part-time people, about $500,000 at an average cost of $50,000 per person. That is well within the per station annual income of between $1 million and $2.8 million we have already described. The remainder would go for supplies, equipment, and land.

Land would have to be leased or purchased for test plantings. It is difficult to be detailed here, as one cannot know without examining a particular area whether a single contiguous farm or plots distributed over various types of terrain and soil would be best. Local farmers might be willing in some cases to lend plots for testing, as is the case in the United States. Local rules and laws on land use, of course, would have to be followed.

The world’s food supply is precarious today, exacerbated by intensifying environmental deterioration. It’s time to add improved perennial crops to the worldwide food and energy agendas.

Recommended reading

G. Conway, The Doubly Green Revolution: Food for All in the Twenty-First Century (Ithaca, NY: Cornell University Press, 1997).

R.H.V Corley and P.B.Tinker, The Oil Palm (Malden, MA: Wiley-Blackwell, 2003).

T.S. Cox, J.D. Glover, D.L. Van Tassel, C.M Cox, L.R. De-Haan, “Prospects for Developing Perennial Grain Crops,” Bioscience 56:649-659 (2006).

L.T. Evans, Feeding the Ten Billion: Plants and Population Growth (Cambridge, MA: Cambridge University Press, 1998).

J.D. Glover and J.P. Reganold, “Perennial Grains: Food Security for the Future,” Issues in Science and Technology, Winter 2010

D. Hillel, Out of the Earth: Civilization and the Life of the Soil (Berkeley, CA: University of California Press, 1991.

W.F. Laurance, L.P. Koh, R. Butler, N.S. Sodhi, C.J.A Brad-shaw, J.D. Neidel et al., “Improving the performance of the Roundtable on Sustainable Palm Oil for nature conservation,” Conservation Biology, 24:377-381 (2010).

R.R.B. Leakey and A.C. Newton, “Tropical Trees: The Potential for Domestication and Rebuilding of Forest Re-sources,” Midlothian, Scotland: Institute of Terrestrial Ecology, Edinburgh Center for Tropical Forests, 1996.

L.H. MacDaniels and A.S. Lieberman, “Tree Crops: A Neglected Source of Food and Forage from Marginal Lands,” Bioscience 29:173-175 (1979).

R.L. Naylor, “Energy and resource constraints on intensive agricultural production,” Annual Review of Energy and the Environment Vol. 21 (1996): 99-123.

S.L. Postel, Pillar of Sand: Can the Irrigation Miracle Last? (New York: W.W. Norton, 1999).

M.W. Rosegrant and S.A Cline, “Global Food Security: Challenges and policies,” Science 302:1917-1919 (2003).

M.W. Rosegrant, X. Cai, and S.A. Cline, World Water and Food to 2025: Dealing with Scarcity. Washington, DC: International Food Policy Research Institute, 2002.

J.R. Smith, Tree Crops, A Permanent Agriculture (New York: Devin-Adair 1953).

D.G. Tilman, J. Hill, and C. Lehman, “Carbon-Negative Bio-fuels from Low-Input High-Diversity Grassland Bio-mass,” Science 314:1598-1600 (2006).

M. Williams, Deforesting the Earth: from Global Prehistory to Global Crisis (Chicago, IL: University of Chicago Press, 2003).


Peter C. Kahn () is professor of biochemistry, Thomas Molnar () is assistant professor of plant biology, and C. Reed Funk () is professor emeritus of plant biology at Rutgers University. Gengyun G. Zhang () is general manager, Department of Agriculture and Bioenergy, Beishan Industrial Zone, Shenzhen, China.

Your participation enriches the conversation

Respond to the ideas raised in this essay by writing to [email protected]. And read what others are saying in our lively Forum section.

Cite this Article

Funk, C. Reed, Gengyun G. Zhang, Peter C. Kahn, and Thomas Molnar. “Investing in Perennial Crops to Sustainably Feed the World.” Issues in Science and Technology 27, no. 4 (Summer 2011).

Vol. XXVII, No. 4, Summer 2011