Nội dung

CHAPTER 10 Natural Systems Agriculture Jon K. Piper CONTENTS Introduction: Environmental Problems Associated with Modern Agriculture The Prairie Model An Agriculture Modeled on the Prairie Ecosystem Elements of the Prairie Model Perennials as Grains Species Diversity Research Agenda and Findings That Support the Model Question 1: Can a Perennial Grain Yield As Well As an Annual Grain? Question 2: Can a Perennial Polyculture Overyield? Question 3: Can a Perennial Polyculture Provide Its Own Nitrogen Fertility? Question 4: Can Perennial Polyculture Manage Weeds, Herbivorous Insects, and Plant Pathogens? Weeds Insect Pests Plant Disease Community Assembly Concluding Remarks Acknowledgment References © 1999 by CRC Press LLC. INTRODUCTION: ENVIRONMENTAL PROBLEMS ASSOCIATED WITH MODERN AGRICULTURE “One Kansas Farmer Feeds 101 People and You,” proclaims a billboard alongside Interstate 135, near Salina, KS. Modern agriculture has been overwhelmingly successful in terms of output per farmer, acre, or hour worked. Agricultural productivity has steadily increased as a result of technological advances in machinery, fertilizer, and pesticides coupled with the intensive use of plant genetic diversity to improve yield through plant breeding. For example, yields of corn and sorghum increased severalfold in the U.S. between the 1930s and the 1980s (Jordan et al., 1986). In terms of return on labor, industrial agriculture based on monocultures of annual grains is unquestionably a highly productive form of food and feed production. This productivity has arisen largely through simplifying agroecosystems into monocultures and tailoring them to maximize yield. In the process, however, many of the links between organisms and the soil that serve to regulate natural communities are ignored or disrupted. The following account surveys some of the more severe environmental effects deriving from large-scale monocultures. With the publication of Silent Spring 35 years ago (Carson, 1962), the public began to become aware of unforeseen environmental consequences of modern agriculture and to question or not whether increasing agricultural production alone was a worthy goal. Three of the most obvious environmental consequences of highproduction agriculture are fossil fuel dependency (and its consequent contribution to global warming), contamination of soil and water with toxic chemical residues, and rates of topsoil loss that exceed the natural rates of soil formation. Additional important consequences include the net depletion of aquifer water for irrigation and the loss of biodiversity from crops, land races, and crop wild relatives. A general consequence of our modern agricultural system is dependency on fossil fuel–based energy. Pimentel et al. (1995) estimate that 10% of all energy used in U.S. agriculture is expended to offset the losses of soil nutrients and water caused by erosion. Over the last few decades, it has taken increasingly more fossil fuel energy to produce a unit of grain in the U.S. (Pimentel, 1984; Cleveland, 1995), with a recent ratio of total energy expended in agriculture (including transportion and processing) to food energy consumed in the U.S. of about 10:1 (Lovins et al., 1995). Another consequence resulting from decades of chemical application on agricultural soils is contamination of surface waters and groundwaters by toxic chemicals. Particularly troublesome are unsafe levels of nitrate derived from applied fertilizer and residues from pesticides aimed at harmful insects, weeds, and pathogenic fungi. Nitrate concentrations in groundwater are strongly correlated with overlying land use (Singh and Sekhon, 1979; Hallberg, 1986). Crops often do not take up all nitrogen applied before it leaches below the zone of biological activity in soil. This excess nitrogen in agricultural soils can slowly leach into deep aquifers, even years after fertilizer application ceases. By the early 1960s, the properties of such long-lived, low-toxicity, and bioaccumulating substances as DDT began to emerge (Carson, 1962). In temperate © 1999 by CRC Press LLC. regions, DDT has a half-life of 59 years. Once bioconcentrated in such top predators as carnivorous fish and bald eagles, DDT sharply reduces the reproductive potential of these species. Long-term exposure has been associated with mutagenesis and carcinogenesis. DDT was banned in 1969; other long-lived pesticides were banned in the 1970s. They were largely replaced with several types of short-lived, acutely toxic compounds (e.g., organophosphates). Residues of many of these subtances are still present all over the planet. Concentration of DDE (a DDT metabolite) is increasing in some Great Lakes and Arctic species (Hileman, 1994). It was not until 1979 that routine agricultural use of the new generation of pesticides was linked to groundwater contamination. Researchers discovered that some of these apparently short-lived, unstable compounds can become extremely persistent once below the soil biologically active zone, where the usual biological degradation does not occur (Zaki et al., 1982; Cohen et al., 1995). Just as with nitrate, complete cessation of pesticide use would not immediately halt the increasing presence of pesticides in groundwater. Many of these chemicals are threats to human health, especially among farmworkers who are exposed directly and rural families dependent upon drinking water from wells. Accumulating epidemiological evidence suggests that agricultural chemicals are associated with increased risks of many types of cancers (Blair et al., 1992; Zahm and Blair, 1992). An additional unintended consequence of widespread and constant pesticide application is evolution of pesticide resistance in target organisms. The result is a need for higher application rates of some pesticides as well as continuous research to develop new substances to control the targeted pests. This phenomenon has been termed the pesticide treadmill; we work harder and harder to stay in the same place but with ever-increasing costs to environmental quality. Despite the enormous problems presented by chemical contamination and fossil fuel dependency, the most serious problem for the long-term sustainability of agriculture is soil loss. Soil erosion is the primary conservation problem on much of U.S. cultivated cropland, and occurs mostly during short intervals of heavy rain or high wind and when the surface is not protected by a mulch or crop canopy (Larson et al., 1997). During the last few decades, about one third of the world arable land area has been lost through soil erosion, and this loss continues at an estimated annual rate exceeding 10 million ha/year (Pimentel et al., 1995). On average, soil on about 90% of U.S. cropland is being lost faster than it is being formed. Because the effects of erosion on some soil physical attributes are irreversible, erosion rates alone may not be good indicators of soil degradation and, consequently, soil quality can decline faster than the erosion rate. Once virgin soil is cultivated, organic carbon rapidly decreases (Campbell and Souster, 1982), large pores crucial for soil function are destroyed, changes in some physical properties increase the rate of erosion, rates of nutrient leaching can increase (Blank and Fosberg, 1989), and populations of such beneficial invertebrates as earthworms decline (Edwards and Lofty, 1975). Prairie soils can lose 30 to 60% of their organic carbon, 30 to 40% of nitrogen, and up to 25% of phosphorus from the A horizon after only a few decades of cultivation (Anderson and Coleman, 1985; Schoenau et al., 1989; Woods, 1989). Many of the © 1999 by CRC Press LLC. consequences of soil degradation, such as reduced crop productivity, have been offset to by improvements in fertilizer and irrigation technology and the development of new, higher-yielding varieties. The industrialization of agriculture, typified by widespread annual monocultures, has led to such profound problems as soil loss, loss of genetic diversity in cultivars, fossil fuel dependency, depletion and contamination of water supplies, pesticide poisoning of farmworkers and nontarget wild species, and development of pesticide resistance in pests. Modern agricultural methods, while highly productive in the short run, are sustainable only as long as topsoil is intact, fossil fuel supplies are affordable, and effective pesticides are available. This may be justified when fossil fuels are cheap and environmental costs can be ignored, but such practices make us vulnerable over the long run. Definitions of sustainability abound. In view of the issues listed above, a sustainable agriculture for the Great Plains should address simultaneously several key environmental problems of modern agriculture. It should feature reduced or eliminated soil erosion, efficient use of land area and soil nutrients, improved water use efficiency, reduced reliance on synthetic nitrogen fertilizer, decreased risk of pest and disease epidemics, effective chemical-free weed management, reduced fossil energy requirements, reduced chemical contamination of soil and water, and the opportunity for farmers to hedge their bets among several agricultural products. A good working definition of sustainable agriculture is one that includes grain production with (1) no chemical contamination of the environment (via pesticides or fertilizers), (2) no dependence on nonrenewables (e.g., fossil fuel, fossil water), and (3) no net soil loss. This working definition is limited in that it leaves out such important considerations as sociology, economy, and justice. But it provides a beginning point for a biological research agenda. By using this definition of sustainability, what type of system could simultaneously satisfy all three criteria? Natural grassland ecosystems provide appropriate models of long-term sustainability because they run on sunlight and rainfall, resist pests, weeds, and disease epidemics, and most importantly because they do not lose soil beyond the natural rate of formation. Some of these aspects are explored in depth in the following sections. THE PRAIRIE MODEL Natural grassland ecosystems may represent our best benchmarks for sustainability. Prairies (1) protect the soil from erosion, (2) provide their own nitrogen fertility requirements through the activities of both free-living and symbiotic nitrogen-fixing organisms, (3) avoid devastation by weedy invaders, insect pests, and plant diseases, and (4) run on sunlight and available precipitation (Table 1). The vegetation structure of prairies has two important general characteristics that contribute to sustainability: the perennial plant growth habit and diversity. Prairies are composed primarily of herbaceous perennial plants growing in diverse arrays. The perennial roots and canopies of prairie plants provide many benefits. These include (1) topsoil protection from wind and water erosion, (2) improved soil quality © 1999 by CRC Press LLC. Table 1 Comparison between Conventional, Industrial Agricultural Systems and Native Prairie Ecosystems for Some Factors that Contribute to Sustainability Factor Fragility Resilience Biodiversity Potential for nutrient loss Connectance (biotic interdependence) Energy sources Nutrient sources Industrial Agriculture Native Prairie High Low Low High Low Low High High Low High Solar, fossil fuel From fertilizers Solar Locally derived, recycled Adapted from J. D. Soule and J. K. Piper, Farming in Nature’s Image, Island Press, Washington, D.C., 1992. With permission. with time, (3) restoration of original soil structure and function following disturbance, (4) biodiversity of soil-dwelling organisms, (5) resistance to weed establishment, and (6) stable populations of beneficial insects. Several studies have demonstrated that the reestablishment of perennial cover on retired cropland can reduce soil erosion while restoring soil quality. The greater root biomass associated with perennial grasses (Richter et al., 1990) gives carbon inputs into the soil that can be several times greater than those into cultivated soils (Anderson and Coleman, 1985; Buyanovsky et al., 1987; McConnell and Quinn, 1988) and reduces rates of nutrient leaching relative to annual crops (Paustian et al., 1990). Active soil organic matter, available nutrients, water-stable aggregates, and polysaccharide content may recover under perennial grasses fairly quickly (Jastrow, 1987; McConnell and Quinn, 1988; Gebhart et al., 1994; Burke et al., 1995). The benefits of a perennial cover were recognized by the authors of the U.S. Conservation Reserve Program (CRP), authorized by Title XII of the 1985 Food Security Act. This program redirected monetary resources and human efforts toward soil conservation and indirectly toward control of agricultural non-point-source pollution. It was designed to protect the most vulnerable U.S. cropland, with a goal to shift 16 to 18 million ha of highly erodible land from annual crop production to perennial vegetation for 10 years. Overall, the program keeps about 595 million t of soil from eroding into U.S. streams and rivers annually, equivalent to a 21% reduction in erosion on cropland (Bjerke, 1991). The second characteristic of prairies that contributes to sustainability is plant biodiversity. Benefits of plant biodiversity include (1) nitrogen supplied by legumes, (2) management of herbivorous insects and some plant diseases, (3) soil biodiversity, and (4) ecosystem stability. Legumes play a critical role in supplying nitrogen to most natural ecosystems. Over periods of several years, perennial legumes can increase the concentrations of both carbon and nitrogen in the soil, as well as influence the size and activity of the microbial community (Berg, 1990; Halvorson et al., 1991). Similarly, legumes are important in providing nitrogen within many pastures as well as multiple cropping systems (Davis et al., 1986; Mallarino et al., 1990). Studies have consistently shown © 1999 by CRC Press LLC. higher dry matter yields in grass/legume mixtures than in grass monocultures (e.g., Barnett and Posler, 1983; Posler et al., 1993). Biodiversity also plays an important role in pest regulation. The presence of nonhost plant species can reduce insect density by interfering chemically or visually with host-finding behavior and thus colonization, feeding efficiency, movement among host individuals, and mate finding (Bach, 1980; Risch, 1981; Andow, 1990; Bottenberg and Irwin, 1992a; Coll and Bottrell, 1994). Moreover, reduced suitability of the microhabitat can reduce insect tenure time, oviposition, and larval survival, and can increase emigration rate (Tukahira and Coaker, 1982; Kareiva, 1985; Elmstrom et al., 1988). In some cases, diverse stands provide a more favorable habitat for parasitoids and predators, leading to reduced levels of insect herbivores (Letourneau and Altieri, 1983; Letourneau, 1987). The weight of published evidence suggests that reduced resource concentration, rather than increased numbers of natural enemies, accounts for most of the observed herbivore reductions within polyculture (Andow, 1991). Similarly, numerous studies have shown benefits of plant species diversity in minimizing certain plant diseases (Burdon, 1987), particularly those diseases vectored by insects (Zitter and Simons, 1980). Establishing host plants within diversified stands can reduce insect landing rate, and thus initial colonization of the plot, by interfering chemically or visually with host-finding behavior (Irwin and Kampmeier, 1989; Bottenberg and Irwin, 1992a,b). This can, in turn, lead to reduced levels of disease in mixtures relative to monoculture (Power, 1987; Allen, 1989; Bottenberg and Irwin, 1992c). Plant species diversity can thus lower the rate of pathogen transmission among individual host plants. Besides these benefits, biodiversity can beget biodiversity. For example, Miller and Jastrow (1993) noted a significant relationship between underground fungal and floristic species richness in prairie restorations. Diversity of soil organisms can have profound effects on plant mycorrhizal associations, nutrient-uptake ability, and nutrient retention and cycling. Biodiversity can have ecosystem-level benefits, too. Several experimental studies have demonstrated that species-rich communities are more resilient and more efficient at using resources than species-poor communities (Naeem et al., 1994). McNaughton (1977; 1985) conducted experiments involving the grasslands of Serengeti National Park, Tanzania. Areas of roughly 16 m2 with different diversities were marked; then exclosures were fenced to prevent grazing by migrant herds of zebra, gazelles, and wildebeest. Grazers reduced the biomass of diverse areas by only about 25%, whereas the less diverse areas lost about 75% of their biomass. Nitrogen limits the number of plant species that coexist in Minnesota grasslands; thus Tilman and Downing (1994) manipulated the number present by varying the amount of nitrogen applied on 207 plots each of 16 m2. They started measuring biomass in 1987. The year 1988 featured a drought which reduced biomass differently on the different plots. After the drought, the species-diverse plots produced half their predrought biomass whereas the species-poor plots produced only about an eighth or less. In these studies, the more diverse communities were more resistant to change. Such studies indicate that plant biodiversity contributes directly to the © 1999 by CRC Press LLC. resilience of grassland communities. Tilman et al. (1996) found that, in experimental plots of perennial grassland plant species, community productivity increased with plant biodiversity. Moreover, soil available nitrogen was used more completely in the more diverse plots, leading to less leaching potential. Finally, resistance to invasion is another collective attribute of complex communities (Case, 1990; Drake, 1991). This property is important for the ability of a community to resist weeds and other exotic organisms. AN AGRICULTURE MODELED ON THE PRAIRIE ECOSYSTEM Elements of the Prairie Model Remnant plant communities of the North American prairie are persistent biotic assemblages in which complex webs of interdependent plants, animals, and microbes garner, retain, and recycle critical nutrients, and protect the soil. As such they provide our best models of the types of communities needed to restore sustainable diversity to compromised ecosystems. Such diverse species assemblages, whose composition varies across soil types, tend to retain species, resist invasion by exotics, and are resilient during short-term climatic variation. Agricultural systems modeled on natural grassland ecosystems would comprise diverse plantings of perennial species that would prevent soil loss, provide much of their own nitrogen requirement via symbiotic nitrogen fixation, and resist invasion by weeds as well as outbreaks of insect pests and plant pathogens. They would be structural and functional analogs of prairie plant communities, composed predominantly of representatives from four major plant guilds: perennial C3 and C4 grasses, nitrogen-fixing species (primarily legumes), and composites (Asteraceae). These functional groups include the majority of prairie vegetation (Kindscher and Wells, 1995; Piper, 1995). A major objective of research toward a sustainable agriculture is to develop innovative methods of production that minimize negative environmental impacts. The CRP, although very successful in terms of soil preservation, is expensive ($1.8 billion/year) (Osborn, 1993) and provides no edible product from the idled land. Hence, the goal of the Land Institute is to develop polycultures of perennial grains that protect the soil and provide the restorative properties of a perennial cover while yielding significant amounts of edible grain. Grain-producing mixtures of perennial grasses, legumes, and composites (e.g., sunflower species) would mimic the vegetation structure and sustainable function of native grasslands in some fundamental ways. Species composition of such perennial mixtures would vary geographically and with soil type. Several promising candidates for a perennial grain agriculture have been identified and evaluated (Wagoner, 1990; Soule and Piper, 1992). Potentially, the sustainable features of such an agriculture include improved soil retention and health, more efficient use of land area, lower fossil fuel dependence, diversity within and between crops to reduce vulnerability to pest and pathogen outbreaks, and greater on-farm predictability and flexibility. Because approximately 20% of U.S. on-farm energy usage is associated with traction (Lovins et al., 1995), perennial © 1999 by CRC Press LLC. grain agriculture, by reducing seedbed preparation and cultivation, application of synthetic chemicals, and irrigation will also translate into savings in energy and materials costs for farmers. Hence, elements of the prairie model to mimic a natural systems agriculture are (1) herbaceous perennials as grains and (2) species grown in diverse fields. The working model, then, comprises several perennial grain species representing four functional groups (i.e., C4 grasses, C3 grasses, nitrogen-fixing legumes, and composites) planted together. To design persistent and diverse prairie-like grain fields, one must be cognizant of the broad similarity of grassland communities as well as the details of their differences. Surveys of locally and regionally occurring species and their local distributions (e.g., Piper, 1995) suggest the types of species likely to participate. First, there are some broad rules that hold across locations (e.g., representation by each of the four major guilds). Second, although one cannot predict perfectly the composition of the final successful community, the general structure of natural communities supported on different soil types gives clues to the types of species to emphasize in the mix (e.g., nitrogen fixers on poor soils). Third, because occasional extreme years can limit diversity considerably (Tilman and El Haddi, 1992), high biodiversity should enable a community to weather better the wide precipitation swings that characterize continental climates. Perennials as Grains The development of perennial seed crops for agronomic mixtures consists of two interrelated efforts. The first is the breeding of new perennial grains. This can involve the domestication of currently wild species as well as the improvement of wide crosses between annual grain crops and their perennial relatives. The second area comprises long-term studies of intercrop compatibility within diversified plantings. This work involves studies to discern beneficial and inhibitory crop interactions, growth and seed yield patterns, and effects on soil quality. Examples of the first approach, the domestication of wild perennials, are experiments with the cool-season grasses mammoth wildrye and intermediate wheatgrass, the warm-season grass eastern gamagrass, the legume Illinois bundleflower, and Maximilian sunflower. Promising examples of the second approach, the development of perennial grains via wide crosses between annual grains and wild congeners, include studies with hybrid grain sorghum (Sorghum bicolor × S. halepense) and “Permontra” hybrid perennial rye (Secale cereale × S. montanum). An obvious consideration before any new crop is adopted is yield. Seed yields of any new perennial grains need to be sufficiently high and stable to make their adoption by farmers compelling. A second consideration is the possible loss of longterm viability as a species is selected for higher seed yield. Studies of several perennial species (Reekie and Bazzaz, 1987; Horvitz and Schemske, 1988; Piper, 1992; Piper and Kulakow, 1994), however, have indicated that, within the ranges investigated, there are no strict “trade-offs” between increased seed yield, vegetative growth, likelihood of future reproduction, or survivorship. Jackson and Dewald (1994) demonstrated conclusively that a severalfold seed yield increase in eastern © 1999 by CRC Press LLC. gamagrass was not associated with a decline in plant growth or survivorship. Similarly, there was no apparent trade-off between higher seed yield and rhizome production in crosses between annual and perennial sorghum species (Piper and Kulakow, 1994). Such results hold promise for research to increase seed yield without losing the perennial nature of a species. Species Diversity Intercropping, the simultaneous raising of different crops in the same place, makes use of species that complement one another spatially or seasonally. Relative to monocultures, intercropped systems can display more efficient use of land, labor, or resources, increased yield, and reduced loss to insects, diseases, and weeds (Francis, 1986; Vandermeer, 1989). Overyielding, a yield advantage in mixture relative to monoculture, can occur when interspecific competition in a mixture is less intense than intraspecific competition or where plant species enhance the growth of one another. Many factors can lead to overyielding. Crops may be released from competition for light by having different light requirements or differences in architecture that minimize shading. Roots of different species may explore different soil layers, or crop species may have complementary nutrient requirements or uptake abilities. Differences in seasonal period of nutrient uptake among crops can also promote overyielding. Intercrops may be more productive than monocultures crops for improving soil fertility, controlling soil erosion, lowering risk of total crop failure, and decreasing crop losses to insects and diseases. Thus, intercropping may satisfy several crop production goals simultaneously. One difficulty facing the plant breeder is that it may be difficult to predict from its performance in monoculture how a crop will behave in polyculture. For example, some plants change their root architecture or patterns of nutrient uptake when grown in association with different species (Goodman and Collison, 1982; Jastrow and Miller, 1993). Shorter plants that are vigorous in monoculture may be shaded out by taller neighbors in polyculture. Thus, selection of varieties for use in perennial polyculture is inherently more complex than selection of varieties for monocultures due to interactions between variety and cropping system that may be unpredictable. So, in addition to the traits needed by any viable crop (e.g., adaptation to the growing environment, tolerance of or resistance to prevailing insect or disease organisms, and reasonably high and stable yield potential), scientists breeding crops for polyculture must also select for or against competitive ability, shade tolerance, and modifications to plant architecture that allow coexistence (Smith and Francis, 1986). Hence, evaluation of species interactions is critical in designing a breeding program for polycultures. Moreover, in perennial systems the outcomes of species interactions may differ in different years (e.g., Barker and Piper, 1995). Mechanisms that reduce overlap in resource demand among coexisting species usually involve differences in location and timing of resource use. Roots of neighboring species may explore different soil layers or a requirement might be met by different resources. Mixtures of legumes with nonlegumes frequently demonstrate yield advantages over monocultures because the two species are tapping different N sources which minimizes competition for this nutrient. Similarly, intercrops may © 1999 by CRC Press LLC. be released from competition for light, and show greater overall productivity, if canopies of component crops occupy different vertical layers (Davis et al., 1984; Clark and Francis, 1985). Differences in length of the growing period or in the seasonal periods of nutrient uptake among crops (e.g., Piper, 1993a) can also reduce direct competition and thus promote overyielding (Francis et al., 1982; Smith and Francis, 1986). Alternatively, a plant may benefit its neighbors by providing cover (Vandermeer, 1980), nitrogen (Wagmare and Singh, 1984), pest or disease protection (Risch et al., 1983; Burdon, 1987), protection from desiccating winds (Radke and Hagstrom, 1976), physical support against lodging (references in Trenbath, 1976), enhancement of mycorrhizal associations (Jastrow and Miller, 1993), or by attracting pollinators (Rathcke, 1984). Indirect facilitation can occur where one species “traps” nutrients that would otherwise leach or be lost from the system, and which later become available to other species (e.g., Agamathu and Broughton, 1985). The Land Institute’s ongoing research agenda in Natural Systems Agriculture revolves around four basic agronomic questions: 1. 2. 3. 4. Can a perennial grain yield as well as an annual grain crop? Can a perennial grain polyculture overyield? Can a perennial grain polyculture provide its own nitrogen fertility? Can a perennial grain polyculture manage weeds, insect pests, and plant pathogens? RESEARCH AGENDA AND FINDINGS THAT SUPPORT THE MODEL Question 1: Can a Perennial Grain Yield As Well As an Annual Grain? Work at the Land Institute to domesticate perennial grains began in 1978 with an inventory of nearly 300 herbaceous perennial species for their suitability to the environment of central Kansas and their promise as seed crops. A second inventory examined the agronomic potential in 4300 accessions of perennial grass species within the C3 genera Bromus, Festuca, Lolium, Agropyron, and Elymus (Leymus). From these inventories, a handful of perennial species was chosen for exploring the principles of perennial grain agriculture. Eastern gamagrass (Tripsacum dactyloides [L.] L.) is a large C4 bunchgrass native from the southeastern U.S. and Great Plains southward to Bolivia and Paraguay (Great Plains Flora Association, 1986). A relative of maize (de Wet and Harlan, 1978), eastern gamagrass has long been recognized as a nutritious and highly productive forage. Because of its high-quality seed (27 to 30% protein and 7% fat, Bargman et al., 1989) and large seed size, however, gamagrass shows much promise also as a grain crop for consumption by people, livestock, or both. Ground seed has baking properties similar to those of cornmeal. The major hurdle facing eastern gamagrass as a grain crop is low seed yield (typically around 100 kg/ha, but as high as 250 to 300 kg/ha in some material at the Land Institute, Piper, unpublished data). © 1999 by CRC Press LLC.