Index | Search | Home | Table of Contents
Myers, R.L. 1996. Amaranth: New crop opportunity. p. 207-220. In: J. Janick (ed.), Progress in new crops. ASHS Press, Alexandria, VA.

Amaranth: New Crop Opportunity

Robert L. Myers

  1. HISTORY
  2. USES
  3. YIELD AND ECONOMIC RETURN
  4. AGRONOMIC RESEARCH
    1. Production Management
    2. Nitrogen Rate
    3. Seeding Rate
    4. Planting Date
    5. Row Width
    6. Crop Rotation
    7. Double Cropping
    8. Intercropping
    9. Cover Crop Effects
    10. Control of Lygus with Alfalfa Strips
  5. PLANT GROWTH AND DEVELOPMENT
    1. Germination
    2. Seedling Vigor
    3. Indicator of Seed Maturity
    4. Response to Moisture Stress
  6. CONCLUSIONS
  7. REFERENCES
  8. Fig. 1
  9. Fig. 2
Grain amaranth (Amaranthus spp.) is a broadleaf pseudocereal (Fig. 1), used by the Aztecs and earlier societies, which began the slow process of coming into modern use and practice in the 1970s. Researcher and consumer interest grew in the 1980s, and formed the basis of continued development of the crop in the 1990s. Consumer interest in amaranth as a food ingredient has been primarily due to the positive nutritional characteristics of the grain, although some of the interest is undoubtedly due to the colorful history of the crop as a "lost" grain. Market demand for amaranth has fluctuated over the last decade, but there has been steady use of the crop for breakfast cereals, snack foods, and more recently, in mass produced multi-grain bread products. Most typically, amaranth products are sold in the health food sections of grocery stores, in specialty food stores, or through direct marketing. Amaranth to-date has only appeared in "mainstream" products when used as a minor component of multigrain foods.

The U.S. has been the leading commercial producer of grain amaranth in recent years, although production has been less than 2000 ha annually since U.S. farmers started growing the crop in the early 1980s. Most U.S. production has been in the upper Midwest and Great Plains, particularly western Nebraska, with widely scattered fields in other parts of the U.S. Amaranth grain species are widely adapted, and have good potential for moisture limited areas. Other species of amaranth have been domesticated primarily for leafy vegetable or ornamental use.

Amaranth's development in the U.S. has benefitted by the formation of private companies that have focused exclusively or in part on amaranth products, and have done much of the work to build public awareness of the crop. The Amaranth Institute, a small non-profit organization of scientists, growers, and agribusiness, has also supported development of the crop through information exchange, promotional activities, and annual meetings.

This paper provides an overview of some of the history, characteristics, uses, and market dynamics for amaranth, and includes a summary of 12 field studies and 2 controlled environment studies with amaranth carried out in Missouri by the author during the period 1990-1994. Information concerning amaranth germplasm and amaranth grain characteristics can be found in previous review papers on amaranth (Kauffman 1992ab; Stallknecht and Schulz-Schaeffer 1993).

HISTORY

Sauer (1993) evaluated the evidence on amaranth's domestication, and concluded that amaranth came into use as a grain at least 6000 years ago in Central America. The oldest seeds found are those of Amaranthus cruentus. Another species, A. hypochondriacus, was in use by at least 1500 years ago, and later became the species used by the Aztecs. Although the earliest seeds of A. cruentus and A. hypochondriacus were both found in caves at Tehuacan, Mexico, Sauer observes they could easily have been domesticated far away, and possibly much earlier than those finds indicate. The origin of the third grain-type species, A. caudatus, is not well understood, but Sauer believes the species was first grown as a grain in the Andean mountain valleys of South America. Amaranth was used as a grain over a wide region, from at least the southwestern U.S., down through Central and South America to Argentina. Amaranth may have had some use as a grain by Native Americans in other parts of the U.S. Although there is no recorded history of its use, A. hypochondriacus seeds found in the Ozarks of the south central U.S. have been dated to 1100 AD.

The most significant historical use of amaranth as a grain crop was in central Mexico during the Aztec civilization (the Aztec name for amaranth was huahtli). Amaranth was both an important food grain for the Aztecs, and a part of their religious practices. Annual grain tributes of amaranth to the Aztec emperor were roughly equal to corn tributes. The reasons for decline in amaranth production following Spanish conquest of the Aztecs in the 1500s are not well understood. Amaranth was significantly used in Aztec religious ceremonies, and some have speculated that the Spanish conquistadors discouraged production of amaranth as part of their overall efforts to suppress the Aztec culture and religion.

Although the production of amaranth in Latin America diminished dramatically subsequent to Spanish rule, the positive attributes of the crop led to its adoption in other areas of the world. By the 1700s, amaranth had spread throughout Europe for use as a herb and ornamental. In the late 1800s, amaranth was reportedly being grown in mountain valleys of Nepal and parts of east Africa. During the 20th century it has been grown in China, India, Africa, and Europe, as well as North and South America.

Although the U.S. has been the leading producer of grain amaranth used in retail food products, the largest production area in the last decade is believed to have been in China (the Chinese amaranth production is reportedly based on cultivars developed by the Rodale Research Center in Pennsylvania). The main Chinese use of amaranth is reportedly to feed the forage to hogs, rather than harvesting the grain.

USES

Although grain amaranth has a variety of potential uses, its only commercial market in the U.S. to date has been for food products. A few small companies have been the primary buyers and processors of amaranth, including Arrowhead Mills (Texas), Health Valley (California), and American Amaranth, now reorganized as Amaranth Resources (Minnesota). These companies sell amaranth both in processed foods, such as cereals, breakfast bars, crackers, and cookies, and also as amaranth flour and whole grain. Beginning in the early 1990s, amaranth also began to be purchased by larger food processors interested in developing multigrain products. Although this trend has led to a broader use of amaranth, market demand has not risen dramatically, due to the small fraction of amaranth used in many commercial multi-grain products, primarily sandwich breads and cereals.

The small but gradually growing market for amaranth as a food grain is based on its nutritional characteristics, and to some extent, its historical interest as the "lost crop of the 'mystic' Aztecs." The nutritional characteristics are indeed positive, with protein content ranging from 12%-17%, and a well balanced amino acid profile, including a relatively high amount of lysine. Amaranth is also high in fiber, and low in saturated fat. The amaranth grain can be popped or flaked, and works well in mixes with flours of other grains, including for extrusion processing.

Although amaranth leaves have been used in both human and livestock diets, little is know about the potential of grain amaranth species for livestock use. When the leaves, stem, and head are used for forage, the product will range from 15% to 24% protein. Palatability is not well known. Within the available amaranth germplasm are genotypes that show potential for significant forage production, but more research in this area is needed.

Amaranth, like most grains, has potential for use in industrial products. The oil fraction of the grain is unusually high in squalene, a chemical that sells for thousands of dollars per ton. However, the percent of squalene in the grain is still small, and may not be economical to extract. Greater promise lies with the starch fraction of the grain. Amaranth, like quinoa (Chenopodium quinoa), has very small, micro-crystalline starch granules, about one-tenth the diameter of maize (Zea maize) starch. The physical characteristics of the starch grains have been cited as being of potential value for both industrial and food product uses, though none has been commercialized to date.

YIELD AND ECONOMIC RETURN

Yields of grain amaranth, like any new crop, are highly variable. Good yields, in relative terms, would be considered to be 1000 kg/ha or better, and these have been achieved in research plots in a number of states throughout the United States (Myers 1994). With a yield of 1000 kg/ha, a farmer with market contracts can easily make a profit, given the typical market price of $0.90 to $1.00/kg for amaranth. However, it is not uncommon for the crop to be a complete failure (little or no stand establishment), or a near complete loss at harvest due to shattering and/or lodging. On the positive side, replicated yields of 3000 kg/ha have been achieved in small research test plots in a few locations. Genetic potential for even higher yields seems to exist, given the productivity of some individual plants. A yield of 3000 kg/ha would be dismal compared to maize, or even sorghum, but for a food crop with a comparable protein level such as wheat, yields of 3000 kg/ha would be respectable, although not outstanding.

Amaranth costs of production are modest, being similar to sorghum. Costs are kept low particularly because fertilizer needs are minimal, and registered pesticides are not available. Although seeding rates are low (1 to 2 kg/ha), seeding costs can be above average (up to $20/acre, or $50/ha) if certified seed is used. Farmers use either certified seed, or, for lower cost, bin run seed. Planting and cultivation costs are equivalent to other grain crops since the same equipment can be used, with some modifications to a planter to handle the small seed size. The other cost that can be high is in the harvest and post-harvest phase. While properly equipped and adjusted grain combines can effectively harvest amaranth, such modifications do add to production costs, from a few hundred to a few thousand dollars. More notably, distance to market is a significant factor for most amaranth producers, since there are only a few amaranth processors and delivery points in the United States (outside of direct marketing). Most growers are faced with shipping the crop hundreds of kilometers by truck, which can lead to a sizable shipping cost for the grain, as opposed to hauling the grain a few kilometers to the local grain elevator. Growers may also experience higher cleaning and handling costs with amaranth than with other grain crops. Altogether, cost of amaranth production per hectare can be expected to be in a range of +/-20% that of maize, with the variability mainly coming from postharvest handling and shipping costs.

The factor that has buoyed the success of amaranth to-date is the relatively high price of the grain ($0.90 to $1.00/kg), which has been up to 10x higher than maize and up to 5x higher than wheat on a per weight basis. Unfortunately, this high price received by farmers has also been a major factor in limiting use of the grain to "health" food products that sell for a premium, or to use as a tiny fraction of a multi-grain mainstream product. Discussions by the author with one major food processor and retailer indicated that for amaranth to gain broader use in the food marketplace, it would have to be sold for about 1/3 to 1/2 of its current price (still a premium compared to other grains), be available in adequate supply (i.e., more production area), and have broader consumer recognition and demand. Certainly amaranth is subject to the same price fluctuations as most grains, when supply overwhelms demand. If amaranth is to ultimately have a much larger and steadier market demand, it will have to follow the path of higher yields and lower prices.

AGRONOMIC RESEARCH

While the United States has led research efforts during the past two decades, grain amaranth has also been receiving research attention in China, India, and Mexico, and by a few scientists in other countries. Most United States scientific work with amaranth during the last two decades has been on production methods, grain characterization and use, and breeding improved cultivars. Although breeding efforts with amaranth in the United States have diminished in recent years, production and utilization research have continued. Cultivar testing has been conducted on a regional basis since the mid-1980s, at over a dozen locations in the United States and Canada. Production research has been conducted in several states, especially Minnesota, North Dakota, and Missouri. Multiple field studies have also been conducted in Colorado, Iowa, Montana, Nebraska, and Pennsylvania. Most of the production research has focused on practical questions such as seeding rates, planting dates, row widths, and fertilizer response. Some evaluation of water use by amaranth has also been conducted in the field. Both insect and disease pests of amaranth have been evaluated.

The predominant source of amaranth cultivars used in the past decade in the United States was the Rodale Research Center in Kutztown, Pennsylvania. Initial breeding was done by Charles Kauffman, with followup work by Leon Weber (Kauffman and Weber 1988). Weber served as the initial distributer of lines for regional testing. Coordination of the regional test was later carried out by Dan Putnam (1990-91), then at the Univ. of Minnesota, followed by R.L. Myers (1992-94) at the Univ. of Missouri. The current regional test coordinator is David Baltensperger at the Univ. of Nebraska. A summary of the regional variety trial results for the period 1985 to 1993 was published in the 1994 annual issue of Legacy, newsletter of the Amaranth Institute (Myers 1994).

Early utilization research with amaranth focused on the grain's nutritional characteristics (Teutonico and Knorr 1985; Breene 1991; Bressani et al. 1992). More recently, work has been conducted on the potential industrial uses of amaranth, and on potential health benefits from amaranth in the diet. The industrial use research has focused on the starch or squalene content of the seeds. The health implications of consuming naturally occurring tocotrienols in amaranth has received preliminary evaluation. Recent food utilization research has included extrusion processing of the grain. Only a few studies have evaluated amaranth cultivars for forage use, and no substantive research has been done on the potential of amaranth grain as a livestock feed.

Production Management

Optimum production practices with amaranth are beginning to be better understood, having been researched in a few different states, particularly North Dakota, Minnesota, and Missouri. What follows are summaries of research studies conducted at Univ. of Missouri during the period 1991-1995, by the author and project staff. Unless otherwise specified, the following management practices were used. Amaranth plots were typically planted in late May or early June, in 76 cm (30 inch) wide rows, at a seeding rate of 2.2 kg/ha. Planting depth was usually 1-2 cm, and planting was done with a 4-row Winter Steiger Precision Plot Planter. Plot size was normally 3 m x 9 m, with each treatment combination replicated four times. Only the middle two rows of the 4-row wide plots were harvested, and plot ends were also trimmed before sampling, all to eliminate border effects. Plots were mechanically harvested with a Winter Steiger Elite Nurserymaster combine. Grain was cleaned, weighed and moisture tested, with recorded yields adjusted to a standard of 10% moisture. Typical data collected besides yield were mature plant height, lodging score, vigor ratings during the season, stand establishment ratings, and as needed, plant population and seed weight at maturity. Occurrence of disease, insects, or other management problems were noted. Amaranth test plots normally were fertilized with 90-100 kg N/ha, with P and K applied according to soil test recommendations for sorghum. Weed control was through pre-plant tillage, row cultivation, hand weeding, and on occasion, Poast herbicide for grass weed control (this is a post-emerge broadcast herbicide legal for research plots, but not labeled for use in commercial fields).

Nitrogen Rate

The effect of nitrogen fertilizer rate on amaranth development and yield was evaluated at two central Missouri sites (Columbia and New Franklin) in 1991 and 1992. The lines used in this study were D136-1, K266, and 'Plainsman'. Rates of 0, 45, 90, 130, and 180 kg N/ha, broadcast preplant as ammonium nitrate, were used. Results indicated that only 45 to 90 kg N/ha were required to reach maximum yield across cultivars, but varieties differed in responsiveness. For the soils tested, it was clear that amaranth does not need as much N fertilizer as maize, or even sorghum. The responsiveness of amaranth to N in Missouri was very comparable to similar research by Elberhri et al. (1993) in Minnesota.

Plant height and lodging were increased by addition of N fertilizer, factors which negatively affected harvesting. Mature plant population and seed weight were unaffected by nitrogen fertilizer rate. Yield compensation was due to higher seed numbers per plant, rather than changes in seed weight, which is a trait that usually held true in other studies on amaranth management (there are differences in seed size between varieties, however).

Seeding Rate

Amaranth response to seeding rate was tested at one central Missouri site (Columbia) in 1991, 1992, and 1993. The lines used in this study were D136-1, K266, and 'Plainsman'. Seeding rates of 0.28, 0.55, 1.1, 2.2, and 4.4 kg/ha were planted (2.2 kg/ha is the most common seeding rate used), all in 76 cm rows. Grain yield was not different for any of the seeding rates. This was due to a combination of two factors, which were the ability of the plants to compensate in seed production per plant at reduced population levels, and also self-thinning of the amaranth stands when planted densely. The lack of response to seeding rates that varied by 16-fold from low to high rate is rather remarkable, given that most crops show a yield response when seeding rate is varied by a factor of 2, and some respond when rates vary by as little as 10% (i.e., maize).

Planting Date

The effect of planting date was studied at one central Missouri site (Columbia) in 1991, 1992, and 1994. The lines used in this study were D136-1, K266, and 'Plainsman'. Three to four planting dates were used each year with about 10-14 days separating each planting (rainfall and soil moisture conditions dictated the dates possible to plant on). Mid-May to mid-June plantings were not different in yield, but planting in early July reduced yield 10%-60%, depending on the cultivar and year of the test. Other studies evaluating amaranth as a double crop (see below) confirmed that planting in early July (after either fallow, or a winter crop like wheat or canola is harvested in late June) reduced amaranth yield. However, such yield reductions also occur with soybeans, used as a double crop on a half-million acres in Missouri. In two studies that did not directly evaluate planting date, amaranth was seeded in the first week of May in central Missouri, with successful stand establishment and good yields. Overall, amaranth has a fairly wide planting window in Missouri, especially if time of maturity and harvest is not a concern. Many northern growers and researchers working with amaranth have tried to time planting so that the crop matures shortly before frost, with a frost serving to accelerate plant dry down and facilitate mechanical harvesting. Missouri tests over 5 planting seasons (1990-94) demonstrated for several lines that amaranth can be successfully harvested without frost, and the plants can dry down at maturity without excessive seed loss, provided no severe storms occur during dry down which exacerbate seed shatter or lodging.

Row Width

Most growers have used wide rows when planting amaranth to allow for mechanical weed control cultivation. However, there has certainly been success with crops such as soybeans in narrow row drilling (albeit in part due to the availability of postemerge herbicides for soybeans). In the early 1990s, no data was available on the response of amaranth to narrow row widths. The effect of row width was evaluated at the Columbia location in central Missouri in 1992 and 1993. Row spacings of 19, 38, and 76 cm were used (76 cm is the standard row width). Narrow row spacing (19 cm) provided good early season weed control, but excessive self competition led to reduced plant height, earlier flowering and maturity, and reduced yield. These responses also occurred, but in a reduced fashion, at the intermediate row width of 38 cm, as compared to the standard with of 76 cm. The excessive plant competition in the narrow row widths was not the result of a higher seeding rate, but reflected the fact that more plants seemed to fully develop in the narrow rows. The reduced self-thinning in narrow rows contrasts with the typical dynamic of wide rows where a smaller percentage of seedlings fully develop, due to self-thinning and in-row plant suppression. This observation is consistent with data reported by Henderson et al. (1993). It might be possible to find a reduced seeding rate that would work combine effectively with a narrow row spacing, but based on the results of these two years of data, narrow rows are not recommended when using a standard seeding rate of 2 kg/ha.

Crop Rotation

A study to evaluate amaranth and canola (Brassica napus) as alternative crops in various rotation combinations with maize (Zea mays), soybeans (Glycine max), and wheat (Triticum aestivum) was begun at Columbia, Missouri, in the fall of 1990. 'Plainsman' amaranth was placed in rotations both as a full season substitute for maize and soybeans, and a double crop substitute for soybeans following wheat. Amaranth was grown in continuous plantings for four years (the studied ended in 1994), in two year rotations with either maize or soybeans, in three year rotations with maize and soybeans, and in a five crop, four year rotation that involved all of the crops above. Amaranth was tested as a double crop after both wheat and canola, in comparison with soybeans (see also double crop study below).

The study had been intended to go for 8 years, but it was terminated after 4 years due to funding cuts, which limited the conclusions that can be drawn. Still, there were some clear findings. For all of the rotations, amaranth had no noticeable allelopathic effects on the following crop, and amaranth crop residue did not present a physical problem for planting and stand establishment of the following crop. Residue levels from amaranth after a winter of decomposition in Missouri are more than soybeans, but less than maize. There did not appear to be any disease or insect accumulation in amaranth grown continuously in this study, but plants were not systematically sampled for disease, nor were insect populations closely tracked. The most obvious problem occurring with continuously planted amaranth was volunteer plants from the previous season(s). Amaranth volunteers are readily controlled in maize and soybeans with tillage and herbicides, but the degree of volunteers in a new amaranth plot led to ragged stands, and also allowed pigweed (a weedy relative of grain amaranth) to build up. Mechanical cultivation to control volunteer grain amaranth and pigweed in a new stand helped, but did not deal with in-row problems. In this rotation study, amaranth planted as a double crop did not perform as well as in a later study that focused on double crop systems.

Double Cropping

From planting date studies, and limited double crop evaluation in the rotation study, it appeared that amaranth had potential as a double crop in Missouri. (In more northern states, amaranth has been strictly limited to use as a full season crop.) To evaluate more fully the potential of amaranth and other alternative double crops, a study was initiated at Columbia (central Missouri) in 1992-93, and expanded in 1993-94 to two additional sites: New Franklin (central Missouri), and Portageville (southeast Missouri) (Pullins 1995). In the fall, large blocks were seeded to canola, wheat, or tilled and left fallow (control). After harvest of the winter crops in mid to late June, the double crops planted were planted no-till, usually by July 1. 'Plainsman' amaranth was planted, along with buckwheat (Fagopyrum esculentum), pearl millet (Pennisetum glaucum), sunflower (Helianthus annuus), and soybean, the standard double crop in Missouri. A partial factorial design was used to also provide comparisons of till and no-till establishment for double crop amaranth, and to evaluate residue effects. Canola and wheat residue were removed from some plots, and applied to selected fallow plots prior to planting the amaranth.

At all three locations, amaranth was successfully established and harvested as a double crop. Amaranth plant height was reduced following winter crops versus fallow, and establishment no-till into wheat residue was difficult, but possible, with the planting equipment available. In two out of the four site-years, amaranth yielded better after fallow than after wheat or canola. Amaranth yields after wheat versus canola were not different in any of the site-years. Double crop amaranth yields ranged from a low of 500-700 kg/ha at Portageville, where lodging was a problem, to a high of 1849 kg/ha following fallow at Columbia in 1993. In two of the four site-years, amaranth yielded 1200 kg/ha or better following wheat and canola, which would provide a substantial profit at current amaranth prices. The top yielding double crop was sunflower, with yields of 1500 to 3000 kg/ha, followed by soybeans, with 600 to 2100 kg/ha. However, amaranth would have provided the greatest profit in this study.

Intercropping

In 1991 and 1992, an intercrop study with amaranth was conducted at Columbia and New Franklin (Clark and Myers, 1994). Amaranth was planted in various row arrangements as an intercrop with cowpea (Vigna unguiculata). The amaranth was compared to pearl millet for intercrop success with cowpea. Row arrangements were alternating rows 38 cm apart, two row strips (on 76 cm spacing), and six row strips (also 76 cm). The cowpeas were planted approximately 10 days before amaranth and millet at each site to allow them to get the cowpea partially established before the faster growing amaranth and millet emerged. In a split application, plots were also treated with broadcast rates of 0, 56, and 112 kg N/ha to gain some information on N dynamics in the intercrop systems.

Amaranth intercropped with cowpea in alternate row fashion had a higher land equivalent ratio (better yield response) than cowpea-millet, but for other row arrangements the intercrop systems were similar. Millet generally outyielded amaranth: amaranth in 2-row strips yielded 973 to 1791 kg/ha, compared to 1750 to 3964 kg/ha for millet; when monocropped, amaranth yields were 1089 to 1591 kg/ha, versus 1849 to 3718 kg/ha for millet. When no nitrogen was applied, amaranth yields were higher in alternate row intercrop with cowpea than when grown as a monocrop, so N and/or other resource complementarity was occurring. An overall conclusion from the study was that amaranth and cowpea can be effectively intercropped, as a variation on the pearl millet and cowpea intercrop system common in parts of Africa.

Cover Crop Effects

Leguminous cover crops have been gaining attention for their ability to help control erosion, reduce weed pressure, and serve as a biological source of N. All these potential benefits are needed in amaranth production, especially since most of the market for amaranth is based on organic production methods. To evaluate the response of amaranth to cover crops, four common cover crops were planted: Austrian winter pea (Pisum sativum), crimson clover (Trifolium incarnatum), hairy vetch (Vicia villosa), and rye (Secale cereale), the last as a non-legume comparison (Fisk 1993). Control plots were kept fallow. Plots received either 0, 45, or 90 kg N/ha, to partially evaluate the N effect of the cover crop. Soil inorganic N was also sampled at three points during the growing season. Austrian winter pea and control (fallow) plots were further split into till and no-till treatments. Cover crops were controlled with paraquat, then residue was mowed down before planting amaranth in early June. Three locations were planted in 1992, the only year of the test, but only two of those were ultimately used for data collection. Those sites, Columbia, and Novelty (northeast Missouri), had much different cover crop growth, since Novelty had a shortage of rain during late spring. Cover crop growth in Columbia was good, but cover crops at Novelty had only modest biomass under the dry conditions.

Overall, Austrian winter pea outperformed the other cover crops, boosting amaranth yields an average of two-fold versus the control plots (from about 600 to 1200 kg/ha). Pea produced the most biomass, and had the highest level of inorganic N during July sampling of the upper soil profile. Amaranth grown after pea and hairy vetch were taller and had higher vigor ratings compared to rye and control plots. On the negative side, lodging was increased in plots following pea. Rye had the expected effect of reducing amaranth vigor compared to control plots. This reduction in vigor was partly, but not completely, offset by the addition of N fertilizer. Surprisingly, amaranth following rye had much higher stand populations in late season than those in control plots or following legumes. It is possible that the reduction in vigor caused by rye also reduced self-competition among amaranth plans, in turn reducing the amount of self-thinning of plants that normally occurs in an amaranth stand.

This study was also the first attempt at planting amaranth no-till in Missouri, an approach later used in double crop studies as well. In this instance, where amaranth was planted following control of growing cover crops by chemical burn down and mowing, no-till establishment worked reasonably well. Seed zone moisture for the amaranth can be helped or hindered by the use of cover crops, depending on rainfall timing preceding amaranth planting.

Control of Lygus with Alfalfa Strips

The most common, and usually most severe, insect pest of grain amaranth is Lygus lineolaris, or tarnished plant bug. This small, brown, sucking insect, slightly smaller than a lady bug, is a general feeder that has many hosts and feeds on a variety of plant parts, but is particularly attracted to the amaranth inflorescence (Wilson 1989). The insect does little damage to amaranth leaves or stems, but can reduce seed weight by up to 70%, and damaged as much as 20%-30% of seeds in Missouri test plots. Worse, the insects damage flowers, and often the vascular tissues feeding a cluster of flowers, preventing the development of sometimes large numbers of flowers into seeds. Pyrethrin based products can provide at least some control of Lygus, but are not normally labeled in a way that allows for use on amaranth.

One non-chemical method of controlling Lygus in cotton has been to plant strips of alfalfa through the cotton fields, using selective mowing to keep parts of the alfalfa in bloom, and thus attracting Lygus. This system of alfalfa strips was duplicated in 1991 and 1992 in amaranth test plots to evaluate its effectiveness at controlling Lygus in amaranth (Clark and Myers 1995). 'Plainsman' amaranth was planted in plots 18 m wide, alternating with 6 m wide strips of alfalfa in a replicated design. Regular sweep net insect surveys were made in the center and edge of alfalfa strips, and at varying distances from the alfalfa into the amaranth strips. Insect counts were made at several times during the season to determine the buildup of Lygus in the two crop strips.

Alfalfa did attract a high percentage of the Lygus relative to amaranth early in the season. However, once amaranth began flowering, the Lygus populations built-up in the amaranth instead, and it was clear the alfalfa did not provide effective control.

PLANT GROWTH AND DEVELOPMENT

Although careful observation has revealed many traits of amaranth's growth and development, there is still much to be learned. In general, a field of amaranth develops the way one would expect for a semi-domesticated crop with little modern plant breeding. Currently available cultivars are highly variable, leading to differences in rate of emergence and growth among neighboring plants, differences in inflorescence morphology and development, and differences in maturity. Amaranth seeds usually germinate quite readily, and when planted shallow (e.g. 1 cm) will emerge within 3 to 4 days in warm soil (³20°C). However, the seedlings are weak, and a crusting soil, heavy rains, or blowing soil particles can easily reduce or prevent stand establishment. During initial development, some plants develop much more rapidly, and vigor differences are accentuated by plant-to-plant competition within a row. Although a mature amaranth canopy is 1 to 2 m in height, individual plants within a row may be much smaller, down to 5 to 10 cm, with tiny grain heads.

When planted in early June in Missouri, inflorescences will begin to emerge by late July, in a pattern that is daylength sensitive for the available cultivars. Anthesis begins to occur after an inflorescence is only partially formed, and continues for parts of the inflorescence until the grain "head" reaches full size (typically 20 to 30 cm top to bottom). Seed development is thus not simultaneous within a head. Seeds that form early begin shattering well before all the seeds on a head are mature, but the seed heads are compact enough to hold most shattered seed in the head, provided strong wind or rain does not occur. (Note: nonshattering germplasm has been identified by D. Brenner, Plant Introduction Station, Iowa State Univ., Ames.) Seeds change in appearance from a glossy, or translucent appearance, to a dull, or opaque appearance as they mature (see research discussion on seed development and maturation).

In states such as Nebraska, North Dakota, and Minnesota, where most amaranth has been commercially grown, amaranth is normally killed by frost before it has a chance to naturally senesce. Plants killed by frost can be harvested a week or so later, depending on drying conditions, which can improve the opportunity to combine the crop before too many seed fall to the ground. However, in more southern areas, such as Missouri, the crop will normally senesce and dry down in Sept., before the first hard frost (in central Missouri, average frost date is end of Oct.). With current cultivars, the plants will turn a medium brown in the grain head and leaves during senescence, then drop almost all leaves. Once senescence begins, stalk strength declines, and plants are more susceptible to stem breakage from high winds.

Because relatively little has been known about many aspects of amaranth growth and development, a series of studies at Univ. of Missouri over a five year period sought to evaluate amaranth's germination requirements, seedling vigor, response to moisture stress, and seed development and maturation. These studies are briefly described below.

Germination

Seed germination tests were conducted in a controlled environment germinator for a variety of light and temperature conditions. Seeds exposed to a 16-h light, 8-h dark cycle, generally had better germination than seeds kept continuously in the dark. To evaluate temperature response, a series of trials was conducted and duplicated at the following day/night temperature regimes: 27°/24°, 24°/21°, 21°/18°, 18°/15°, 15°/12°, and 12°/9°C. Germination percentage and speed of germination declined when temperature was reduced to the 21°/18°C regime, and germination percentage was greatly reduced at 18°/15°C. or below. Amaranth seeds normally sprout within 24 hours at warmer temperatures, and the hypocotyl elongates enough within 72 hours for emergence from a 1 cm depth, which corresponds to typical field emergence of 3-4 days at shallow (1-1.5 cm) planting depths.

Seedling Vigor

It has been repeatedly noted that amaranth seedlings grow at different rates, and that in mature stands, plants vary from 5 to 200 cm in height. It has also seemed that some seedlings are eliminated through self-thinning. A simple study was designed In an attempt to determine if these vigor differences are genetic, competition induced, or both. Seeds were planted in bedding plant flats, with each unit approximately 3 x 5 cm across and 5 cm deep, with about 1 cm spacing between adjacent units in a flat. Either 1 or 3 seeds was planted in each unit, with some flats having all units seeded, and others having only alternate units seeded. In this way a simple comparison of seed competition could be done, both with seedling roots interlacing within a unit, and with physical barriers between individual seedlings. Also, above ground competition could be compared based on whether adjacent units were planted or empty. Results indicated both genetic variability for seed vigor, and competition effects when plants were as little as 3 cm tall. The taller the seedlings grew (they were observed until the tallest were about 10 cm in height), the more pronounced the competition effects became relative to the genetic differences. Number of leaves was slightly reduced in plants that were being outcompeted. What is unusual about amaranth, compared to more established grain crops, is the inconsistent pattern of seedling vigor. Most grain crops, when planted densely, would have all plants more or less equally reduced in vigor, but with amaranth, some plants just "outrun" the others, suggesting amaranth cultivars are not homogeneous.

Indicator of Seed Maturity

In most research plots and commercial fields of amaranth grown in the 1980s, harvest timing was simply tied to the first hard frost that killed the plants, allowing 5-7 days for dry down, and then combining the crop. In a more southern region such as Missouri, however, amaranth cultivars such as 'Plainsman' often dry down before frost, making timing of harvest more important. Although mechanical harvesting is easiest when plants are almost completely dry, it is likely seed shattering will be significant by that time point. Also, the dry down period for grain amaranth in Missouri was sometimes up to 3-4 weeks. It was observed in 1990 field plots that amaranth seeds seemed to turn from translucent to opaque as the plant matured. To evaluate if the change in seed appearance corresponded to physiological maturity, seed samples were collected in a fashion intended to evaluate the relationship of seed coat appearance with their seed maturity.

In 1991, heads were collected and threshed from replicated plots near Columbia, Missouri. Comparisons were made of seed appearance (translucent vs. opaque), seed weight (fresh and dry), and seed germination, for three cultivars ('Plainsman', D136, and K256), two planting dates (June 3 and 21), and three sampling dates. The samples were taken at approximately two week intervals. In 1992, a more limited sampling was done of 'Plainsman', D136, and 'Amont'.

Results indicated the percent of opaque seeds increases as the plant ages. The dry weight of the opaque seeds was higher than translucent seeds, and opaque seeds were generally lower in percent seed moisture at time of sampling. Based on this data, it was clear that opaqueness in amaranth seed is a suitable indicator of maturity. However, opaqueness was not an exact measure of physiological maturity (physiological maturity is defined as the point when seeds reach maximum dry weight, and generally coincides with the seeds being developed enough to germinate). Opaque seeds did have a higher percentage of germination than translucent seeds, but some seeds rated as translucent were still able to germinate.

Other characteristics of opaque seeds were noted which distinguished them from translucent seeds. Opaque seeds are harder and more brittle. It was also noted, under microscopic examination of seed cross sections, that opaque seeds have a perisperm with a white, fine granular appearance inside, as opposed to the glossy, more globular perisperm appearance of translucent seeds.

Response to Moisture Stress

Amaranth has often been described in the literature as a drought tolerant crop, but the veracity of this claim had not been validated by any scientific tests. Two field studies involving comparative analysis of multiple grain crops, both alternative and traditional, provided some insights into the relative drought tolerance of amaranth, and the crop's physiological response to moisture limited conditions. One study evaluated several alternative and traditional grain crops in soils with varying depths of flood-deposited sand, while the other study evaluated the same set of crops in constructed field plots of varying soil depth.

Response to sandy soils. Eight different grain plots were planted into replicated plots in the Missouri River bottoms near Jefferson City in 1994. The intent of this field test was to evaluate the relative performance of traditional and alternative grain crops on newly created sandy soils versus 'regular' alluvial soil in the river valley. The sand deposits were the result of record flooding along the Missouri river in 1993, which covered approximately 150,000 ha of bottomland fields with 15 cm or more of sand. A site was selected that had sand ranging in depth from 30 cm to 120+ cm.. Two adjacent field areas were used, one with "shallow" sand (30-45 cm) and one with "deep" sand (>75 cm). A nearby plot of ground with no sand deposits was chosen as a check plot area, and designated as "regular" soil (silty loam topsoil, formed as floodplain alluvium).

The eight crops chosen were soybeans and sorghum as traditional crops, and sunflowers, pearl millet, cowpeas, mung beans (Vigna radiata), amaranth, and sesame (Sesamum indicum) as alternative crops. Crops were selected on the basis of their suitability for planting in early summer, and expected tolerance for moisture limited conditions created by sandy soil conditions.

Stand establishment of all eight crops was good in the regular soil plot area. In the sandy plots, establishment was delayed by dry conditions until the first rainfall, which came a week after planting. Only amaranth germinated prior to rainfall, which was rather remarkable considering the very low moisture content in the seed zone of the sand. After rainfall, emergence was slow for some crops due to inconsistent planting depth in the sand. Ultimately, good stands were obtained of all crops except sunflowers and amaranth, which had partial stands. The partial stands of amaranth were probably due to the very dry seed zone conditions when the seed was planted, and began to germinate. Although some amaranth seedlings became established, other seeds probably only imbibed water, or barely sprouted, before running out of moisture. The other crop seeds which sat idle until the first rainfall a week after planting, had much more moisture available during establishment. The tendency of amaranth seeds to imbibe moisture and start germinating even under rather limited moisture conditions has been noted in regular field plots, and is somewhat of a problem, occasionally leading to ragged stands, where some plants emerge quickly, and others emerge later after rainfall, or not at all.

It was expected that tap-rooted crops like sunflower and amaranth would do well under the soil conditions of this study, where flood-deposited sand was layered over silt loam soil. However, both sunflower and amaranth tap-roots were found to go only as deep as the wetting front in the sand. Relatively modest rains of 1 cm or less would wet the sand down to about 10-15 cm. The tap-roots would go that deep, then turn sharply at a right angle (running parallel to the surface) as if they had hit an impermeable barrier. The relative advantage of the pearl millet and sorghum (grass) crops was their ability to send out a mass of fibrous roots laterally that could capture a larger portion of the rainfall that had occurred. The tap rooted crops (sunflowers and amaranth) would likely have been much more successful if early rainfall had wetted the sand all the way through to the underlying silt loam, allowing the tap-roots to reach that depth. It was noted that weedy amaranths (pigweed), established themselves quite successfully from rainfalls that occurred during the month preceding the test planting on June 14 of amaranth and the other grains. Pigweed was the dominant weed by far in the newly deposited sand areas, which overall were surprising free of weeds in the first year after flooding (weed seed probably floated off while the sand dropped out of flood waters as a heavy material; the sand was deepest near levy breaks).

Response to soil rooting depth. In 1994, a study evaluating response to rooting depth was planted at the Univ. of Missouri Agronomy Research Center, Columbia. The same 8 crops as in the sand test above, plus foxtail millet (Setaria italica), were planted into a unique constructed plot area that had been created in a previous field test. The site consisted of 6 x 60 m soil pits running parallel to each other, and separated by about 4 m. The pits had been dug out to varying depths and bottom-lined with a heavy grade, impermeable black plastic. On top of the plastic was laid a small drain tile. The pit was then filled with top soil to the surface level. Thus, the deepest pit was filled with about 94 cm of topsoil, which is certainly a better root zone through that depth than the regular soil profile of that site, with topsoil down to only about 20 cm, and somewhat of a claypan underneath. Not surprisingly, some of the crops planted on the 94 cm deep plot (thus with 94 cm of topsoil) actually had better vigor and more height than those planted on the regular soil, at least until moisture stress became limiting in early Aug.

The crops were planted in replicated 3 m wide strips across the field perpendicular to the direction of the soil pits, creating plot units 3 x 6 m in size. The soil depths varied from 33 to 94 cm (Fig. 2).

The rainfall pattern in May through August at the test location was conducive to gaining some valuable insights into the response of these crops to moisture deficiency. During May and early June, there was adequate rainfall to establish the crops to a height of 30-60 cm. From late June through early Aug., there was a dry spell broken only by a couple of minor rainfalls of about 2 mm each. During this dry spell, crops began to wilt, first in the shallow plots, then successively in the deeper and deeper plots marching across the field. By early Aug., after almost no rain for almost 6 weeks, even the crops on 94 cm of topsoil were affected, some severely. By contrast the crops on the undisturbed soil showed no wilting or other obvious adverse affects from the dry conditions.

The progressive wilting across the field in successively deeper soils was no surprise, but provided a chance to observe wilting patterns in some crops, including amaranth, for which little information is available. What was surprising was the way amaranth dealt with moisture stress compared to eight other grains. Before the test, it was presumed that amaranth would not fair particularly well in shallow soils, based on the idea that amaranth is tap-rooted like sunflower (sunflower gets its drought tolerance by sending its tap-root down deeply, as opposed to being efficient in water use). It was a complete surprise that amaranth was the first of the nine crops to show wilting symptoms. The wilting pattern of amaranth was quite dramatic, with plants going from full turgor one day to completely limp (leaves hanging straight down) the next. However, after almost 10 days of the plants in the shallow plots being completely wilted, the amaranth plants made a stunning recovery to full turgor and apparent health (no leaf lesions or brown leaf edges) after a small rainfall of 2 mm. This pattern repeated itself in deeper plots, after further dry conditions, and another 2 mm rainfall several days later. In contrast to amaranth's ability to rebound so completely from wilting, sunflower had substantial leaf tissue death, as did other crops, even though they were wilted for fewer days than amaranth, and perked up to some extent after the small rains. These observations indicated amaranth may owe part of its reputed drought tolerance to an ability to shut down transpiration through wilting, then recovering easily when moisture is available.

Amaranth was not noticeably affected by moisture limitation in the deepest plot (94 cm), unlike soybean, which used up the moisture available in the profile and completely died by mid-Aug. (soybeans on adjacent regular soil were seemingly unaffected by the dry conditions). Sunflower wilted severely in the 94 cm plots before rain finally came in mid-Aug. In other words, amaranth ran out of moisture first in 33 cm of soil, but for some reason was better able to get by on the moisture in 94 cm of soil than crops such as sunflowers and soybeans. This would seem to indicate amaranth has some water use efficiency compared to those other crops, at least during later stages of growth.

There is still much to be learned about amaranth's water use patterns, especially if the crop is to be promoted for moisture-limited agricultural areas around the world. The only other significant evaluation of amaranth water use in the United States was conducted by Henderson et al. (1992) in North Dakota. Neutron probes were used to evaluate seasonal water use in four grain amaranth cultivars. Data indicated amaranth water use extended down to about 1.2 m under their soil conditions (Prosper, North Dakota), with no cultivar differences in water use. Total water use was typically about 27-32 cm in 1990-91, but was half of that during 1992, which was an unusually cool season with limited plant growth.

CONCLUSIONS

The current status of amaranth is as a crop which has great potential, a variety of possible uses, and a decade-plus of research behind it. However, as with most alternative crops, cultivar improvements are needed, production and utilization research challenges remain, and major barriers exist in market development. More specifically, further breeding should improve seedling vigor, reduce lodging and seed shatter, improve ease of harvesting, and improve yield. Insect pests need to be better understood, and control methods developed. Likewise, disease symptoms and prevention are poorly understood with amaranth. Although information on amaranth response to basic management practices had been developed, including in the studies reported on in this paper, additional work could be done, on factors such as pH, NPK interaction, effects of soil type and seedbed preparation methods. Amaranth response to moisture stress deserves more study. From a utilization standpoint, a fair amount has been learned about general amaranth grain characteristics. However, the response of factors such as grain protein to management and environment is only partially understood. Also, specific high value uses of amaranth, such as the starch fraction, merit additional investigation.

Markets remain relatively small and undeveloped, in part because there is a general lack of familiarity with amaranth in the public and private sector. To achieve a higher level of market penetration, amaranth will have to become more publicized, prices will have to fall (although a premium could still be commanded), and availability will have to be increased. Distance to buyers is a problem for many current amaranth growers. Special markets such as the starches or other seed components could lead to increased marketing opportunities.

On the positive side, amaranth is widely adapted, tolerant of dry conditions, and diverse germplasm is available for breeding to improve the crop. Amaranth has relatively good yield potential for a high protein grain crop, especially considering the lack of breeding with the crop. It can be grown successfully with conventional grain crop equipment, usually with only minor modifications, and has a production cost comparable to other grain crops. The colorful appearance of the crop and its colorful history continue to generate interest in the crop, and its good nutritional characteristics combined with its variety of potential uses illustrate the importance of continued work with this "rediscovered" crop.

REFERENCES

Fig. 1. Grain amaranth research plots at Colombia, Missouri. Grain heads, or inflorescences, are fully developed at this stage of the season, in late Aug.

Fig. 2. Cross section diagram showing how restricted root zones of varying depths were developed by digging pits or trenches across a field, with the bottoms lined with impermeable plastic, and then backfilled with topsoil. Pits ran 60 m east-west, and were 6 m wide. Crops were planted north-south perpendicular to the length of the restricted root zone areas. In the undisturbed soil, a claypan is prevalent in the B horizon.

Last update August 15, 1997 aw