Accelerating Research and Development
The natural interactions of an ecosystem supplemented with biological-control organisms, products, narrow-spectrum synthetic pesticides, and resistant cultivars form the basis of the EBPM systems proposed in this report. Implementation of these new pest control systems will require more knowledge about ecosystems than does the use of conventional pest management systems. Because the knowledge base in this area is limited, considerably more research will be required to develop and implement EBPM. Acquiring that knowledge base will necessitate coordinated efforts among many scientific disciplines.
The development of organic pesticides was a direct consequence of the enormous investment in organic chemistry research by national governments and industry since the late 1800s, beginning with dye chemistry research by German industry. Unfortunately, there is currently no such comprehensive investment in the kinds of research needed to help direct development of the various components of EBPM. EBPM research will benefit from the biotechnology knowledge base generated primarily through health science research; but a comparable knowledge base in ecology, particularly microbial ecology, is needed.
A full understanding of the dynamics among pests and control organisms in agricultural and forest ecosystems will require development of in situ methods to measure populations of both pests and control organisms and characterize their interactions—from multitrophic effects to molecular signaling. Managed ecosystems are complex and vary spatially and temporally. Pest and control organisms evolve and migrate; populations vary from field to field and among plants in an individual field. Diagnostic tools to measure pest populations must be useful both to individual growers and to researchers investigating pest management effects on a larger geographical scale.
Investigations will provide an understanding of the basis for the stability of these communities in natural systems as well as identify where the use of supplemental inputs and cultural practices disturbs the managed ecosystem and how pest populations develop and adapt to these disturbances. By more fully developing the ecological knowledge base and then coupling that with the expanding base of pest management experience, scientists can devise strategies to effectively manage pests and restore balance to forest and agricultural ecosystems. For example, research investigations of chemical signaling between insects and plants and production of toxic proteins by bacteria have been instrumental in developing biological-control products such as semiochemicals and biological insecticides.
Research is also needed to ensure that EBPM strategies can be transferred from the laboratory to a grower's field where agronomic practices and biological-control inputs can be evaluated in whole-farm settings. Moving from discovery to implementation requires a systems approach in which researchers from many disciplines cooperate in building safe, profitable, and durable pest management approaches.
FOUNDATIONS OF A KNOWLEDGE BASE
Currently EBPM research is the primary focus of a relatively small group of scientists whose contributions have led to the development of successful and economically feasible pest management systems. The challenge now is to move beyond optimal examples and into the mainstream of pest management. Accelerating the development of EBPM requires a clear agenda and institutions that can carry out that agenda. The focus should be to identify research that either overcomes obstacles to wider use of ecologically based management systems or leads to new and innovative ecologically based approaches.
Thus far only cursory knowledge about biological factors that control agricultural pest populations has been developed. For example, it is known that release of biological-control organisms to control pests is an option; however, the ability to predict the outcome of that release is limited. In fact, the proportion of releases of classical biological-control organisms that have resulted in complete suppression of the pest is rather low (Hall and Ehler, 1979; Hall et al., 1980).
EBPM implementation requires a basic understanding of the ecosystems in which agricultural predator and prey coexist, including an understanding of
the many interacting factors that influence population size and the activities of control organisms and antagonists,
the composition and dynamics of microbial communities present in soil and on plant surfaces,
the food webs governing the population of pests, and
the factors influencing the spatial distribution and differential fitness between weed and crop.
A knowledge-based approach to pest management requires an understanding of these factors, communities, and trophic levels—i.e., the interactions among organisms within the managed ecosystem. To predict an outcome of these interactions requires knowledge of
vulnerable stages in organismal life cycles,
factors influencing both pest and biological-control organism reproduction,
disease vectors, especially vectors of viruses, and
molecular signals governing pest and biological-control organism interactions.
It may not always be possible to find ''natural" tools and tactics to control pests, neither will "natural" approaches always be the most ecologically sound. Many natural products are toxic to a broad spectrum of beneficial organisms and to humans; such products include neem, a pesticide extracted from the neem tree, and pyrethrum, an insecticide found in dried flowers of several Old World chrysanthemums. These natural compounds need replacement as much as do broad-spectrum synthetic chemicals. Because durability is an important goal of EBPM, organisms and chemicals that are specifically targeted to affect pest species but relatively benign to all other organisms must become the dominant management tools used in agricultural and forest systems of the future.
PRIORITY RESEARCH AREAS
A national research agenda, both general and cross-cutting, should identify broad areas of ecological research that promise to yield the critical information needed to accelerate progress in EBPM. The eight broad categories of research priorities identified by the committee can be used as a guide to the process of understanding how this complex, evolving system can be managed to reduce pests. These areas should receive priority for funding:
research on the ecology of managed ecosystems;
research on behavioral, physiological, and molecular mechanisms to effect EBPM;
research to identify and conserve natural resources necessary for EBPM;
development of better research and diagnostic techniques;
development of ecologically based crop protection strategies;
research on implementation and evaluation of EBPM;
research to improve understanding of the socioeconomic issues affecting adoption of EBPM; and
development of new institutional approaches to encourage the necessary interdisciplinary cooperation.
Research on the Ecology of Managed Ecosystems
Knowledge of managed ecosystems must guide the development of pest management tools. In the past 20 years progress has been made in answering questions about the ecology of managed systems; however, answers to old questions have generated new questions and challenges. The committee suggests that research in agricultural ecology over the next 10 years should focus on the areas noted below.
The Stability of Communities Containing Pests and the Communities' Sensitivity to Perturbation
Probably the most important research needed to learn how to manage pests is research to understand the basis for stability of plant, arthropod, and microbial communities. Soil-inhabiting pests and plant-parasitic nematodes present special problems because most species of soil microbial communities have not yet been catalogued. Because many species that form ecosystem communities are not known, community dynamics and effects on soil pathogens and nematodes are not well understood. In many cases, it is not known whether ecological theory developed for macroorganisms also applies to microbes. A recent treatise on the subject declares that "microbial ecology is experimentation in search of theory" (Andrews, 1991).
To the extent they can be characterized, microbial communities appear to be altered only transiently in response to the introduction of biological-control organisms. The basis of this stability in community structure is not likely to be understood until the major components of the community have been identified and their roles elucidated. With the identification of the components of soil communities, and their functions, new organisms may be identified that, on augmentation, could serve as effective biological-control organisms.
Most studies of nonpest soil invertebrates have focused on determining how the diversity and abundance of species is affected by specific changes in cropping systems (e.g., tillage and pesticide use), but very little is known about how the sometimes significant changes in the species composition of these communities affects survival and growth of pest populations. Studying organisms in the soil milieu has been extremely difficult using conventional methodologies, highlighting the need to develop new tools for these investigations.
Physical, Chemical, and Biological Conditions in the Rhizosphere and the Phyllosphere and Their Effect on the Development of Pest Populations
Soil conditions affected by cultural practices and chemical inputs (e.g., fertilizers and pesticides) in turn affect the activities of microorganisms, arthropods, and weeds. Unfortunately, it is not known which specific soil characteristics
have the most significant impact on these organisms. For example, it is known that green manures enhance general microbial activity in soils planted to potatoes and result in the suppression of Verticillium wilt (Davis et al., 1994; Schroth et al., 1992); however, the specific mechanisms by which this process occurs are poorly understood.
More is known about the components of microbial communities in the phyllosphere than in the rhizosphere, yet there is still much to learn about the conditions that affect the microbial populations in the phyllosphere. On leaf surfaces, the population density of bacterial foliar pathogens is known to undergo striking changes in response to the environment (Hirano and Upper, 1990), but little is known about how population dynamics of such pathogens affect other components of the phyllosphere community or vice versa. Competition is thought to play an important role in determining the dynamics of microbial populations in both the rhizosphere and phyllosphere (Loper and Lindow, 1993), but science has only hints as to the nature of this competition.
Attempts have been made to reduce frost damage caused by bacteria that nucleate ice on plants by introducing strains of that same bacterium that lack the gene responsible for the ice-nucleating activity. Such competition experiments with bacteria, engineered using recombinant DNA technology to remove the ice-nucleating genes, were the first reported experiences in which recombinant organisms were released in the environment (Lindow, 1993). These experiments focused public concern on issues related to the use of genetically engineered organisms to solve agricultural problems. Although these experiments served to draw attention to the issue of risk associated with biological control, they also showed the potential of this new technology to build an increased understanding of microbial ecology.
The Role of Dispersal Mechanisms in Establishment of Populations of Major Pests and Their Biological-Control Organisms
Many agriculturally important weeds, invertebrates, and microbes are not permanent residents of the crop fields and orchards they inhabit. Because of annual (or longer term) changes in the suitability of specific fields or geographic regions for survival of these species, many persist based on their ability to physically shift the location of their populations. The movement involved in these shifts can be limited to a few kilometers or range from the southern to northern boundaries of the United States. Some of this movement involves establishment of only a small number of individuals to initiate populations in new locations. In other cases, however, millions of microbial spores or swarms of arthropod pests can invade an area within a few days and cause rapid destruction of a crop.
Although progress has been made in understanding the movement of many key pests, information is still needed to accurately predict the timing and direction of that movement. Such predictive power can provide farmers with ad-
vanced warning of invasions, which in turn may persuade many growers to use responsive instead of prescriptive pest management methods. Furthermore, a more complete understanding of movement of beneficial organisms can lead scientists to develop methods to promote efficient movement of these organisms between heterogeneous agricultural habitats and between agricultural and nonagricultural areas.
Inoculation of Agricultural Systems with Specific Biological-Control Agents: Effects on Nontarget Organisms
There has been much concern about the nontarget effects of conventional pesticides; but as biological organisms, products, and resistant cultivars become the centerpiece of pest management, more attention will need to be focused on their potential side effects. When an introduced biological-control organism fails, it will be necessary to determine why it failed, rather than simply search for alternative agents. It will also become extremely important to know whether an organism slated for introduction has the potential for long-lasting negative effects.
There has been justifiable concern about introduction and release of non-indigenous and genetically engineered biological-control agents. There are many unanswered questions about the ecological effects of these organisms; and in most cases the techniques or tools to properly obtain this knowledge do not exist. Thus it will be necessary for ecologists and molecular biologists to conduct interactive interdisciplinary research to develop new tools for studying the biology of beneficial organisms. In situ hybridization and new methods for analysis of DNA extracted directly from environmental samples are expected to provide further insight into the distributions and populations of beneficial microbes.
Manipulating the Spatial and Temporal Structure of Agricultural Habitats to Reduce Pest Outbreaks
An agricultural habitat's suitability for survival of invertebrates and microbes varies greatly over both time and space. Change over time is especially pronounced in temperate zones where agricultural crops are rotated on a seasonal basis; the change in mean temperature combined with the change in crop species effects a dramatic, wide-scale habitat alteration for most organisms. Agricultural habitats also can vary in suitability on the scale of meters and centimeters. Plant pathologists and entomologists have long known that outbreaks of some pests are directly related to the spatial layout of crop fields as well as the location of fallow land. Weed scientists who have studied the spread of nonindigenous weed species know that land use patterns influence invasion and establishment of these species. Landscape ecology, which took shape as a formal subdiscipline of ecology in the early 1970s, focuses specifically on the response of individual
plant and animal species and communities to distinctive landscape patterns. Geographic information systems (GIS) and remote sensing technologies—new tools in this area of study—are leading to rapid advances in understanding how temporal and spatial variation affect survival of endangered species, pest populations, and beneficial organisms.
One technique of landscape ecology used by weed scientists involves manipulating spatial access to water and nutrients in order to suppress weed populations and increase crop yield. Thus, techniques and tools developed by landscape ecology scientists and traditional techniques in life-system ecology can be used to identify subtle changes in some agricultural landscapes that could improve the competitive advantage of beneficial organisms while negatively impacting pests. The effects of specific changes in the agricultural landscape on the overall durability of any local cropping system can only be adequately addressed by interdisciplinary teams including weed scientists, plant pathologists, entomologists, biochemists, economists, sociologists, crop scientists, etc.
Biological Characteristics that Enable Organisms to Adapt to Changes in Physical, Chemical, and Biotic Conditions
It is known that repeated use of pesticides conditions many targeted pests to naturally select for resistant biotypes. Currently, the arthropod class contains the largest number of species to have developed resistance to pesticides. However, weeds developing resistance to herbicides is a growing problem in crop production; in fact, researchers have discovered plant pathogens that can overcome resistant cultivars within just a few growing seasons. Although the mechanisms and dynamics of pesticide resistance have received increased attention over the past 40 years, more research is needed to understand how both detrimental and beneficial agricultural organisms adapt to their environment. The little information that does exist on this topic indicates that organisms in agricultural and forest systems readily adapt to the many abiotic and biotic changes in their habitats.
Classical tools of ecological genetics and population genetics, and new tools of molecular biology, are being used by researchers in attempts to slow the rate at which pests adapt to chemical and physical stresses and, at the same time, increase the rate at which beneficial predators adapt to these stresses and to the defenses of their target prey. For example, Lewis and Martin (1990) found that preconditioned parasitoids are more effective than nonconditioned parasitoids when released into the environment.
Better Ways to Understand How Multispecies and Multitrophic-Level Interactions Contribute to Pest Population Dynamics
Plant pests are components of ecological systems. Their presence is made possible by many ecological factors, and their effects result in multiple ecological
interactions. Concerns related to multispecies and multitrophic interactions are leading to some experimentation and field applications.
Plant pathologists are infecting plants with certain minor pathogens that protect the plants from more devastating pathogens.
Weed scientists are investigating how the interactions of a crop infested with multiple species of weeds affect yield.
Research teams are discovering higher concentrations of beneficial organism attractants in wild plants than in their commercial counterparts.
The knowledge base in this area has expanded in the past decade, but a complete understanding of such multitrophic-level interactions is essential for designing effective and reliable biological-control measures for EBPM.
Research on Behavioral, Physiological, and Molecular Mechanisms to Effect EBPM
The mechanisms that govern interactions among organisms have been an area of intense study in both medical and agricultural research because understanding these mechanisms is key to understanding how pests and diseases can be controlled through active intervention. For agriculture, one of the pay-offs of such knowledge would be the ability to quickly develop genetically engineered plants resistant to pests. Already the toxins responsible for the success of Bacillus thuringiensis (Bt) as a biological-control agent offer promise for the enhancement of plant resistance to specific arthropods when the genes encoding the toxins are incorporated into plant genomes.
Knowing what factors control interactions among organisms will help scientists better understand how to manipulate pest populations, or communities containing these organisms, to reduce their impact. Understanding biological control at the cellular or molecular level has a predictive value, enabling scientists to set realistic expectations for when and where biological control may succeed or fail. For example, if biological control is known to be mediated by iron competition between an antagonist and a target pathogen, it is not likely to succeed in conditions in which iron is replete. Identification of the characteristics that determine biological-control activity is prerequisite to manipulating these characteristics through genetics, modification of behavior or the environment, or other interventions.
The discussion of research needs in the following section provides suggestions as to how additional knowledge of processes directly related to biological control can be used to enhance biological-control processes. The committee's recommendations are meant to focus on gaps in current research priorities that are uniquely related to biological-control issues. Although there have been some major developments in studies of plant-host interactions (e.g., the finding that a specific plant host, by cellular signaling, indicates recognition of a pathogen),
such issues as arthropod signaling, antagonism/toxicity/ antibiosis, host selection, target-host diseases, target-host parasitism, and plant-disease resistance are subjects in need of understanding at behavioral, molecular, and physiological levels.
Manipulating the Behavior of Pests to Reduce the Risk of Pest Outbreaks
Modifying pest behavior to reduce pest damage is a more long-lasting approach to pest management than pest destruction; however, success with behavior modification of pests and natural control agents has been restricted to a limited number of cases. There have, however, been some advances made in molecular biology and behavioral studies that have highlighted the need for researchers to cooperate on signaling system projects. Tumlinson and his colleagues formed an interdisciplinary team to investigate plant, herbivore, and parasitoid interactions (Tumlinson et al., 1993). They found that oral secretions of herbivores activate plants to emit a set of volatile signals that benefit both the plant and parasitoids by alerting the parasitoids to the presence and location of their herbivore hosts. They further report that these biological-control organisms learn to process visual and olfactory information to more efficiently locate plant pests.
Although "behavior" is generally considered to be a property of multicellular organisms, microbes also respond to signals that control behaviors such as mating, "feeding," movement, and germination; so behavior-based management techniques could also be developed for some microbes.
Knowledge that some arthropod and nematode pests must locate and recognize a crop as a food source can guide scientists to chemicals capable of controlling pest foraging behaviors. An understanding of organisms' behavior in managed ecosystems can also help scientists design cropping systems that favor the success of biological-control organisms and reduce pest damage (Kareiva, 1990; Lewis et al., 1990). For example, arthropod pests can be lured away from a valuable crop by planting a less vulnerable or less valuable trap crop nearby. Knowledge gained from manipulating the behavior of arthropod pests could be useful in research to manage weeds and pathogens.
Signaling between Target Hosts and Biological-Control Organisms
There is much interest in identifying signals and elucidating the nature of signal reception and transduction in organisms. Chemical signals influence an organism's growth and development and affect interactions between different organisms, such as in the induction of host-defense responses. Much is known about the role of signal molecules and physical cues in determining the behavior of arthropods; but more comprehensive knowledge of the nature and function of these signals is needed if they are to be exploited for biological control of pests,
and investigations should address the behavior of the pests in situ, not just in the laboratory.
One role of signaling between different organisms within communities can be seen in the response of pests to signals from potential host plants. For example, acetosyringone released from wounds of certain plant species induces the transcription of the vir genes of Ti plasmid-containing strains of Agrobacterium tumefaciens, initiating the process of crown gall formation (Gelvin, 1992). It is reasonable to assume that similar signaling occurs between many of the coevolved components of rhizosphere and phylloplane communities.
One of the most useful applications of chemical signaling has been the use of pheromones for trapping and quantifying populations of specific arthropods. Although there has as yet been no comparable practical applications in the control of plant pathogens, pheromones are known to control fungal sexual cycles (Spellig et al., 1994). With better understanding, fungal pheromones can be used to control fungal behavior.
Using Natural Antagonism, Toxicity, and Antibiosis to Control Pests
During the past decade, compelling evidence of the ecological roles of natural antibiotics produced in situ by microorganisms has been reported (Weller and Thomashow, 1993). Whereas many scientists formerly believed that antibiotic compounds were produced as artifacts of laboratory culture conditions, it is now known that such compounds can be and are produced by microorganisms—e.g., B. thuringiensis and A. radiobacter—inhabiting natural substrates and in concentrations adequate to inhibit pests. Microorganisms that produce antibiotic compounds are among the most successful biological-control agents. Future research could lead to
production of novel antibiotics through domain switching, as has been done for polyketide antibiotics;
understanding when genes encoding antibiotic biosynthetic enzymes are expressed in situ and what environmental cues induce antibiotic biosynthesis;
optimizing in situ antibiotic production by placing biosynthetic operons under regulatory control of constitutive promoters or by modifying the biological environment to enhance biosynthetic gene expression; and
evaluation of pest resistance to antibiotics or toxins produced by biological-control agents.
To date, manipulation of genes encoding the -endotoxin (Bt toxin) of B. thuringiensis has received the most attention for practical application as a biological insecticide. B. thuringiensis produces massive amounts of these crystalline insecticidal proteins, which differ in shape, size, and host-specificity. Bt-toxin genes are being introduced into plants as resistance genes; into endophytic bacteria that reside in the xylem of plants; and into bacteria that, after heat-
killing, serve to encapsulate the toxin, thereby enhancing its persistence in the field. Some products using these strategies are currently commercially available, others are in the development stage. The success of Bt-toxin research raises optimism that other organisms are also potential sources of effective biological-control products.
Using Pathogens of Plant Pests for Biological Control
Pathogens of diseases that afflict pests offer potential for use as biological controls. Unfortunately, the time lag between when the disease is introduced and when it becomes established in the pest population limits this method; crop damage during the incubation period often is too extensive to be economically acceptable. This is the problem with baculoviruses, which otherwise offer considerable potential for biological control because of their specificity and safety. Research to identify the initial steps of baculovirus pathogenesis can lead to the development of fast-acting baculoviruses. Other research is being directed at developing disabled ("suicide") recombinant baculoviruses which would eliminate only the arthropod that received the virus. Disabled viruses are attractive because there is only a remote possibility that a foreign gene introduced into the viral genome by recombinant DNA technology could enter the ecosystem. Research plans are also being developed to determine whether some microbial pathogens that have coevolved with arthropods could be manipulated to control specific arthropod pest populations.
A virus of the plant pathogenic fungus Cryphonectria parasitica is an important model to describe how naturally occurring or recombinant viruses can control fungal diseases of plants (Nuss, 1992). Work with this biological-control agent has shown the potential for use of viruses to control fungal diseases. Although use of viruses of fungal pathogens could result in a reduced virulence of the fungal pathogen population, much more information about viruses is needed before their potential for biological control can be exploited fully.
Perhaps the greatest potential for the use of diseases as biological controls is in weed management. Research is needed to enhance or conserve naturally occurring soil microbes that have weed suppressive abilities. Unfortunately weed management via microbial agents will not be viable until it is possible to (a) increase the speed and effectiveness of the control agent in the field and (b) ensure that the agent will not damage beneficial plants. Understanding the genetic basis of pathogenicity, virulence, and host-range of microbes that attack weeds could result in the selection and development of safer and more effective microbial weed-control agents.
Improving Strains of the Parasites, Parasitoids, and Predators of Plant Pests
To date, the major emphasis in genetic improvement of arthropod predators
Biological Control of Chestnut Blight
Will blight end the chestnut?
The farmers rather guess not.
It keeps smoldering at the roots
And sending up new shoots
Till another parasite
Shall come to end the blight.
The American chestnut tree once comprised more than 25 percent of the eastern hardwood forest. It persists today in its native range only as sprouts or as decaying remnants of the past. Soon after the discovery in 1904 of Cryphonectria parasitica on an imported Oriental chestnut in New York, the imminent demise of the American chestnut from chestnut blight was well recognized. In the 1930s, the pathogen was observed in Europe, where the European chestnut also succumbed to chestnut blight.
Approximately two decades after chestnut blight was first diagnosed in Europe, certain trees in northern Italy appeared to recover from the disease. Isolates of C. parasitica obtained from healed cankers were hypovirulent: they infected European chestnut but rarely produced lethal infections. Later, stands of American chestnut that survived blight were found in Michigan. Many isolates of C. parasitica obtained from healed cankers of Michigan trees also were hypovirulent. Hypovirulence is caused by a double-stranded RNA (dsRNA) virus that infects the fungus, reducing its virulence on chestnut trees.
When hyphae of a hypovirulent and virulent isolate of C. parasitica fuse, the virus and other cellular contents can be transferred from the hypovirulent to the virulent isolate. On acquisition of the virus, the virulent isolate becomes hypovirulent. In chestnut trees, transfer of the virus from the hypovirulent to the virulent form occurs by this process and, as a result of viral infection of the virulent fungus, the tree can actually recover from chestnut blight. Dissemination of the dsRNA virus among isolates of C. parasitica can occur naturally; natural dissemination restored the chestnut as a dominant forest species in northern Italy. Naturally recovering stands in Michigan are also testaments to the success of transmissible hypovirulence for control of chestnut blight.
Despite the presence of hypovirulent strains in the hardwood forests of the eastern United States, natural recovery of the chestnut forest has been slow and geographically limited. Fusion of hyphae, and subsequent transmission of the dsRNA virus, can occur only between isolates of C. parasitica within a single group (called an anastomosis group). The population of C. parasitica is very diverse in the eastern United States, and the presence of multiple anastomosis groups is one factor slowing the dissemination of hypovirulence to native virulent isolates in this region. Recently, a transgenic isolate of C. parasitica containing the viral genome was constructed. Transfer of the viral genome from the transgenic isolate to other isolates of C. parasitica does not depend exclusively on hyphal fusion. Therefore, the transgenic isolate may be an effective source of the virus for transmission to virulent isolates of C. parasitica in the eastern United States. Because C. parasitica does not infect the root systems but only the shoots of the chestnut tree, the genetic diversity of the eastern chestnut forest has been preserved throughout this century.
SOURCE: Nuss, D. L. 1992. Biological control of chestnut blight: An example of virus-mediated attenuation of fungal pathogenesis. Microbiol. Rev. 56:561-576.
and parasitoids has been on developing natural control agents that are resistant to pesticides so that they might function within a system dominated by broad-spectrum pesticides. In the future, both classical breeding techniques and genetic engineering could be used to increase the efficiency of biological-control agents in specific agricultural systems. Characteristics that could be altered include heat and cold tolerance, host range, and migratory cues.
Certain fungi trap and feed on plant-parasitic nematodes. These and fungi that parasitize other fungi, such as Trichoderma spp. and Gliocladium spp., are useful for biological control of soil-borne diseases in greenhouses (Lumsden et al., 1993). More can and should be done to determine the importance of parasitism in the biological-control activity of these fungi and to characterize the molecular basis of parasitism. As with other biological-control agents, more information is needed about their inherent genetic variability as well as information about which characteristics of the biological-control organism are important targets for strain improvement.
The Basis of Host Selection and Host-Range Specificity
An important concern of the general public about the release of non-indigenous predators and parasitoids is their potential to harm indigenous nonpest species. To allay these concerns requires a better understanding of the genetics of host-range specificity, which will provide insights into how organisms can be genetically altered to limit their current and potential host range and thus enhance their safety as control organisms.
Some progress has been made in understanding the genetic basis of host selection by microbial plant pathogens—an important step toward understanding the molecular basis of host specialization by plant pathogens. In fungal, bacterial, and viral plant pathogens, genes that determine the spectrum of plant cultivars that pathogens can infect have been identified. Staskawicz and colleagues (1984) changed the host specificity of a selected pathogen by deleting or inactivating a single gene. The host range of a plant pathogen was narrowed genetically by Mellano and Cooksey (1988).
The genetic basis for host selection and host specificity in arthropod predators and parasitoids is less understood than it is for plant pathogens, but progress is being made in this area. It is known, for instance, that host specificity of an arthropod parasitoid is determined by a virus that the parasitoid transmits to its host when it inserts one of its eggs into the host. It is also known that within a given host range, arthropod parasitoids can adjust their host preferences by learning cues that lead to a suitable host. Although predators with a broad host range have preferences for specific prey, the genetic factors that govern preference are not understood. Research in the area of host selection could have long-term benefits in improving the efficiency and safety of these biological-control organisms.
Pest-Resistance Mechanisms of Plants
Host plant resistance is the most important component of many plant disease management systems. The development of molecular biology research tools has resulted in identification and analysis of genes that control the interactions of plants and pathogens and is contributing to efforts to produce transgenic crop plants resistant to a variety of plant pests.
Although resistance is less central to management of arthropod pests than of diseases, it is a major component of certain control programs such as the wheathessian fly and small grains-greenbug systems. Such programs may become more important for all crops when research on genetic engineering of arthropod-resistant plants gains momentum. Research has focused on expressing toxic proteins in crop plants to kill pest species. Scientists are now limited to use of a few toxic proteins such as Bt toxins, but this may change based on (a) screening of novel microbes for active compounds and (b) research into physiology and biochemistry of metabolic and hormonal systems unique to certain taxa of arthropod pests. The recent discovery of an insecticidal cholesterol oxidase produced by a fungus offers the possibility that other such toxins will be discovered. The fact that the activity of Bt toxins and other protein-based toxins is limited to arthropods is encouraging. Industry is involved in discovering novel toxins, and academic research in arthropod physiology and biochemistry is likely to help identify arthropod-specific targets and novel approaches for engineering plants.
Plants can respond to infection by an active resistance response, termed induced resistance, which is maintained systemically throughout the plant for a period of time after the infection. This response can protect the plant from some subsequent infections by plant pathogens. Recent progress in understanding the mechanism of signaling the advent of such challenges between different parts of the plant offers hope that this phenomenon will soon be explained at the molecular level (Kessmann et al., 1994). An understanding of the process could lead to novel, biologically based means to control damage caused by some pests. Clearly, basic studies in plant biology can have important impacts in agriculture. Continued support of research in this area is required if the full potential of these and other approaches is to be realized (National Research Council, 1989b).
Research to Identify and Conserve Natural Resources Needed for EBPM
A founding principle of classical biological control of exotic pests is that natural enemies of the pest can be found in the geographic region where the pest evolved. Likewise, plant geneticists have found that the best place to find pest-resistant plant varieties is in the geographic region where the plant and pest coevolved. Collection and identification of germplasm from such regions has been a high priority for plant breeders. It is probable that every pest has at least one biological-control agent that could be identified by this approach.
Systematic exploration for biological-control agents, however, has been limited to a relatively small number of arthropod and weed pests. For fungi, bacteria, and nematodes, even less exploration has occurred. Research to identify, collect, and document the natural enemies of pests in their native habitat should continue to be an important component of efforts to discover useful biological-control agents.
Preserving biodiversity is justified by the benefits to be accrued from identifying beneficial products, such as taxol; and the economic benefits of preserving potential biological-control agents as part of biodiversity are expected to be significant. A classic example is the discovery of the pesticidal properties of the neem plant; its extract acts as an ''antifeedant," eliminating the urge to eat for a variety of arthropod pests. Random screening for such chemicals would probably not be cost effective; a logical and more cost-effective alternative would be enlisting the aid of indigenous people knowledgeable about the regional flora to identify unique biotic products that have known pest-control properties (Thurston, 1990).
Systematics Research and Taxonomic Resources
Reliance on conventional chemical pesticides for suppression of agricultural pests has obscured the need for continued development of systematics or taxonomic resources; however, careful characterization of pest populations is important to the success of the more target-specific methods of EBPM. Taxonomists are continuously challenged to accurately identify an increasingly diverse group of pests and potential biological-control organisms that includes both arthropods and plant pathogens. Unfortunately, systematics research has not kept pace with the growing demand, and many of the most important groups of pests and biological-control agents are poorly characterized. Even in agricultural systems that have been studied for decades, many natural enemies of agricultural pests do not fit into described species and therefore their classification remains uncertain. In unmanaged ecosystems where effective natural control agents are found in association with their hosts, the number of undescribed species is even higher.
Revised classification at the family, genus, and species levels is needed for a broad array of important pest and antagonist groups. The present taxonomy of microorganisms is oriented to medical clinical isolates, a situation that presents a special problem for biological control as regards regulation. Because the taxa are structured to characterize clinical isolates, many isolates from nature that cause no known clinical problems are often labeled as pathogens because there is no related nonpathogenic taxon in which the isolate can be placed. The problem has been exacerbated by the widespread use of commercial data base systems that are almost entirely focused on identification of clinical bacteria. Accurate identification of organisms used by pest management researchers depends on an understanding of the systematics of the organism and on the availability of tools to
make accurate nominative determinations. Unfortunately, revised classification at the family and genus levels may require 10 to 20 years to complete, and few such studies have reliable financial support.
Because populations of pests within a species are heterogenous, characterization of the organism is, in many cases, needed at the subspecies or "strain" level, where conventional alpha taxonomy has not provided sufficient support. Molecular, numerical, and other advanced biological tools are now available to address some of the heretofore intractable problems in systematics such as strain characterization and species discrimination. Tools including DNA amplification, DNA sequence analysis, hybridization studies, restriction fragment length polymorphisms, and various advanced forms of electrophoresis are increasingly used in research laboratories. Using these tools could provide the taxonomic foundation essential for development of EBPM.
Development of Better Research and Diagnostic Techniques
Researchers should devise new methods to study, monitor, and evaluate agricultural and forest system processes as well as the effectiveness of pest management tools. Providing researchers and farmers with improved diagnostic techniques is key to both developing and increasing the use of biological organisms, products, and resistant plants. The examples provided here are not comprehensive, but illustrate how development of methods and techniques are needed to understand and implement EBPM strategies.
Methods for in Situ Study of Microbial Populations
The lack of methods for studying the growth and development of microbes in situ hampers implementation of microbes as biological-control agents. Although it is recognized that only a small percentage of microbes present in natural habitats can be cultured (Oliver, 1993), current methods rely almost exclusively on culturing techniques to isolate, enumerate, and characterize microbial control agents and target plant pathogenic bacteria and fungi. Novel approaches based on in situ hybridization, immunoisolation using monoclonal antibodies, or amplification of DNA isolated directly from environmental samples offer promise for the in situ study and quantification of microbial populations. Future studies using such techniques will greatly enhance understanding of the composition of communities of plant-associated microbes and population sizes of introduced antagonists or weed or insect pathogens as well as to increase understanding of the complex interaction between soil microbes, allelopathy, and resulting weed-suppressive soils.
A further constraint to the implementation of microbes as control agents is the inability to evaluate their in situ activities, including the production of critical metabolites or expression of phenotypes essential for biological-control activity.
Many phenotypes that contribute to the biological-control activity of such microbes can be readily detected and quantified in culture but are difficult to assess in situ with current methods. The development of reporter gene systems (i.e., a gene encoding a phenotype that is readily detected and quantified in natural habitats serves as a "reporter" of the activity of a gene that is critical to biological-control activity) offers promise for the assessment of in situ activities of microbial control agents (Lindow, 1995). Creative development and application of new methods for the detection and characterization of microbes in situ must be encouraged.
A significant limitation to progress both in understanding biological-control processes and in developing better biological-control agents is the lack of effective methods to genetically manipulate these organisms. Understanding has advanced based on scientists' use of model systems. Unfortunately, the tools developed to genetically manipulate these model systems often do not work with organisms of interest for biological control. Transformation methods and vectors need to be developed for organisms potentially important for biological control.
Grower-Friendly Diagnostic and Monitoring Methods
EBPM will require a substantial input of information at the farm level, and its success will depend on accessibility of input procedures. The development of diagnostic kits that can be conveniently used and reliably interpreted for measuring and monitoring populations of pests and natural enemies, detecting weed seeds, and assessing pest contamination of planting material are examples of areas in which research is needed to develop usable procedures.
Production, Stabilization, and Delivery of Biological-Control Organisms and Products
An important challenge to implementation of EBPM will be to provide reliable sources of living biological-control organisms. Because most of this information is maintained as carefully guarded trade secrets, it is difficult to assess the status of efforts developed to maintain the viability of living biological-control organisms and products in commercial formulations. There is more of a history of private-public sector cooperative research in fermentation or rearing of biological-control agents, but it is still largely funded by industry. Opportunities for cooperative research exist also in developing methods to diminish production of phytotoxic or otherwise deleterious metabolites of microorganisms and in identifying optimal physiological conditions for the storage of biological-control agents. For example, it is thought that Gram-negative bacteria, which lack the resistant spore state of many Gram-positive bacteria and fungi, cannot withstand formulation and have a short shelf life. There has been, however, research indicating that
the "stationary phase" may be a stress-resistant phase of the Gram-negative life cycle (Kolter et al., 1993) that is analogous in many respects to spores of Gram-positive bacteria. Future research may be directed on genetic selection of biological-control bacteria with traits for stress resistance.
Another obstacle to implementation of biological control is the large amount of inoculum required for efficacy of certain microbial agents. Research to optimize placement of biological-control agents in close proximity to target pests and production of propagules at the most effective physiological state can do much to remedy this problem. These are the types of issues that could be fruitfully explored through cooperative private-public sector research efforts.
Development of Ecologically Based Crop Protection Strategies
Previous research priorities have addressed components of various biological-control strategies, however, the goal of durable pest control can only be achieved if the various strategies are coordinated and implemented in an integrated manner. The concept of IPM provided a good model for the type of integration that would be desirable, but IPM also serves to exemplify pitfalls to be avoided. For EBPM to succeed, there must be strong input from scientists involved in all pest-related disciplines.
Because pest management strategies are linked to other farming system components, it will be necessary to study relationships among, for instance, pest management, plant nutrition, and farming practices. By doing so, researchers will gain comprehensive data—essential for the development of an integrated crop protection strategy. The focus of this section, however, is directed toward EBPM-related research.
Develop Predictive Models for Cropping Systems
Integrating biological and cultural (physical) methods of pest and disease control is key to achieving long-term pest management. The biological/natural concept for managing pests has historically been ecologically based. Such an approach recognizes that managed cropping systems exist within and are a part of an ecological web of relationships where physical and biological processes are interactive and dynamic. The plants, herbivores, predators, pathogens, weeds, etc., that exist within this web have developed a repertoire of offensive and defensive maneuvers and countermaneuvers in response to one another. At present, the primary limitation to durable cropping systems is insufficient knowledge of the relative importance of system interactions often in and of themselves as well as in relation to the system as a whole.
Science needs to be able to predict the consequences of the various potential interactions within a single cropping system. Individual interactions must be understood, and attempts to link the various biotic/biotic and biotic/abiotic pro-
cesses into a predictive system must be made. Integration will be a challenge both because of the complexity of the problem and because of the requirement that different disciplines must coordinate efforts and work together. Models can provide a theoretical basis to predict the interactions of the various biotic and abiotic components of an ecosystem. However, the knowledge gained must be applicable to agricultural and forest production systems.
The type of research needed is much like that addressed in IPM studies. For EBPM, however, biological-control practices must be integrated with each other and with normal crop-management practices. Such integration will require a variety of investigations:
timing of applications,
methods of application,
amount of biological control applied, and
whether controls can be applied simultaneously and what would be the outcome.
For example, although there is well-justified interest in developing microbial mixtures for biological control, it has been observed that certain strains degrade the antibiotics produced by a second strain when the two strains are mixed together. In this case, the combination of the two strains may be less effective than if each strain were applied individually. An integrated scientific approach can direct research efforts toward practical pest-management solutions.
Develop Strategies for Achieving Durable Plant-Host Resistance
Concern has been expressed about the durability of pest-resistance genes, but pathogen-and arthropod-resistant cultivars can be deployed in ways that reduce or eliminate the breakdown of resistance caused by changes in the genetics of pest populations. Innovations in genetics and applied evolutionary biology provide an excellent opportunity for researchers to develop methods to augment the few commercial practices available for protection of pest-resistance genes. There is a need to increase scientific collaboration between plant molecular biologists, pest molecular biologists, evolutionary biologists, and crop scientists so that real progress can be made.
The durability of resistance genes has long been an area of active research that has led to strategies of deployment of multigenic resistance, multiline varieties, and varietal mixtures. The dangers inherent in the genetic vulnerability of crop plants will need to guide both research and commercial management of resistance genes. This will be particularly true of resistant genes introduced into transgenic plants because they may be widely distributed among plant taxa. With the discovery and use of these genes should come studies investigating ways to increase their stability after deployment.
It is also important that assessment of resistance genes not be limited to only
direct effects on plant pests. Resistance genes may also affect natural enemies of the pests and microbial communities in the rhizosphere or phyllosphere. In the manipulation of plant genetics, thought should be given to how traits of crop plants can be altered to enhance the overall health by enhancing populations or activities of naturally occurring or introduced biological-control organisms.
Research on Implementation and Evaluation of EBPM
The research needs for EBPM discussed in the sections above have emphasized the need for new knowledge of biological processes, interactions, or organisms useful for biological control. Emphasis should be placed on implementation research to increase adoption of EBPM. There is a need for applied implementation research, as well as farm-scale and area-wide evaluation of the biological and socioeconomic impacts of new management tactics.
Large-scale field trials are important to gaining grower acceptance of EBPM. In a survey of entomologists in the 12 north-central states, the lack of funding of farm-scale research was identified as one of the major constraints to the implementation of biological controls in pest management (Mahr, 1991). A large-scale trial can be performed on an operational farm where researchers can document actual grower decision making in the context of ecosystem processes and agronomic practices (National Research Council, 1989b; Shennan et al., 1991). To gain additional information, researchers can design complementary experiments to control these variable agronomic practices (Shennan et al., 1991). The complex interactions among farming system components warrants further investigation of pest-management implementation in whole-farming systems.
Moving from the discovery phase to the implementation phase requires information on scale-up, compatibility of new and established technologies, and other factors that effect the viability of approaches in commercial agriculture. Some components of private industry can efficiently move new products or procedures from small-scale tests to farm-level demonstrations of efficacy. Small industries, however, such as certain producers of predatory and parasitic arthropods and microbials, do not have the resources to conduct implementation research. Furthermore, pest-control practices that are not product-related do not have private-sector support, although funding for the discovery of these practices through competitive funding agencies is often available. Investments in discovery research should be accompanied by investments in the implementation of new discoveries. Increased attention to this area of research should result not only in increased implementation, but also in directions for discovery research.
Evaluation of the Impact of New Technologies
It can be argued that a concerted evaluation of the early impact of the chemical-control paradigm on factors such as agricultural and economic long-term viability, environmental contamination, and human health would have provided warnings of potential problems. Implementation of EBPM strategies must be tied to studies of the effect these new methods may have on nontarget species, agricultural practices, and human health.
Research to Improve Understanding of the Socioeconomic Issues Affecting Adoption
Social and economic factors will play significant roles in determining whether EBPM will be widely adopted. Research is needed to (a) develop management systems that are both economically and technically feasible, (b) target research and development at problems where solutions have the greatest value, (c) develop methods that can prolong the economic life of EBPM, and (d) provide the foundation for public policies that facilitate the development and implementation of EBPM.
The social and economic research needed can be divided into three broad categories:
methods for measuring the direct and indirect effects of pest management systems for use in benefit-cost accounting;
interdisciplinary research to compare the economic benefits and biological performances of different EBPM systems; and
clarification of factors that influence risk-taking by agricultural producers.
Methods for Measuring Direct and Indirect Effects
It is now apparent that benefit-cost analyses of EBPM systems, including their initial and long-term social benefits and costs, should be factored into setting priorities and developing public policies for agricultural production. Such analyses will allow a rigorous comparison of alternatives and provide decision makers with the full range of costs and benefits of EBPM. It also will provide a systematic way of comparing one approach to another as well as provide a common framework for the analysis of alternatives.
Benefit-cost analyses should include an account of both direct and indirect costs and benefits of alternative management systems. Some direct benefits and costs will involve comparative cost of pesticide replacement versus the cost of the biological-control agent, the application costs, and effects of cultural practices that may be necessary to enhance the effectiveness of the treatment. Indirect benefits and costs involve
human health effects to workers and consumers;
environmental effects of the two practices on soil, water, and air quality;
effects on nonpests, including wildlife and beneficial organisms;
limitations on land use, such as reduced flexibility in crop rotation and crop selection;
establishment of other pests;
development of resistance by pests to the control procedures; and
lack of reliability of the management practices.
Although the direct costs of alternative agriculture systems may be greater than a chemically based system, indirect costs of the alternative system may be relatively low. Research is needed to identify the appropriate variables in comparing pest management approaches, and methods must be developed to collect comparative data. This will involve integrating the work of natural and social scientists, two groups that traditionally have not worked together. Such cooperation between disciplines, however, is important to developing cost-benefit analyses because many of the benefits of EBPM are indirect but essential to responsible ecosystem management. At the current level of understanding, if all indirect costs and benefits were considered in decisions concerning adoption of pest-control practices, benefit-cost analysis probably would indicate a significant advantage for EBPM.
Research to Compare Economic Benefits and Biological Performance
Economic feasibility is the most important factor to producers considering alternative management systems. Unfortunately, economic feasibility studies of the systems comprising EBPM are limited. Studies that have been done are often too narrow, ignoring decisions at the grower level and the impact of commodity prices, or estimating the profitability of the ecologically based management system without considering the profitability of traditional systems (Reichelderfer, 1981). Studies of the economic feasibility of ecologically based management systems relative to conventional systems need to be done for a variety of crops and geographic areas. Understanding the relative economic performance of this type of management system is essential to adoption and will focus future research and development on which aspects of the management system should be optimized.
The lack of systematically collected economic and performance data comparing the various pest-control methods continues to limit cost-benefit analyses. In 1979, because data on the degree of pest suppression and yield for the evaluated practices were not available, Reichelderfer (1979) was forced to conclude that all the pest-control methods evaluated achieved the same degree of pest control and produced the same crop yield. Similar problems with the lack of data constrained evaluation of different control methods for soybean cyst nema-
tode (Reis et al., 1983) and the navel orangeworm, Amyelois transitella (Headly, 1983).
Clarification of Factors that Influence Risk Taking
Risk, and the way producers manage it, is an important factor in determining the speed of adoption of new technologies. Few studies have attempted to measure or otherwise quantify growers' attitudes toward the risk of using EBPM; however, Antle (1987) showed that the attitudes of populations and individuals toward risk can be estimated. Knowing the distribution of risk attitudes in a population of producers could help refine strategies for how new pest management systems should be presented.
Development of New Institutional Approaches to Encourage the Necessary Interdisciplinary Cooperation
The difficulty of manipulating the ecology of a cropping system to reduce losses caused by pests must not be underestimated. Every crop faces multiple assaults from numerous arthropods, diseases, and weeds, simultaneously; thus management decisions become complex. Many of the chemical and cultural tactics (e.g., tillage and crop rotation) available to agricultural and forest producers have broad impacts on the system. In suppressing one pest, it is likely that other components of the system will be disrupted. Research into the interactions among organisms in managed systems will enable scientists to design tools and methodologies to reduce populations of problematic organisms without negatively affecting the balance of the system.
The complexity of managed ecosystems necessitates coordinated multidisciplinary and interdisciplinary research to develop and implement EBPM. Funding for single-investigator initiated research projects will continue to be important, but for the most part single-investigator research will not sufficiently address the multidisciplinary nature of biological-control research. It is, therefore, essential that research institutions, and research policy, facilitate multidisciplinary and cooperative research. Frequently institutions, traditions, and policies erect barriers to interdisciplinary research. Efforts must be made to explore the nature of these barriers and to pull them down.
Despite the generally recognized need for a systems approach to solving pest problems, the operating philosophy and organizational infrastructure in pest management is largely fragmented between different disciplines, agencies, and research institutions. This fragmentation is partly a result of the tendency for science to solve a problem by breaking it into parts--i.e., disciplines. Science
education and research experiences are largely clustered around subject disciplines. For example, a biologist may specialize in molecular and cellular processes, or in whole organisms and organism-organism interactions, or in analyzing patterns that govern whole ecosystems. It is rare that a single researcher is accomplished in research across all these levels of biological organization. Research expertise can be further segmented by types of organisms and functions such as virology, entomology, plant pathology, microbiology, plant physiology, plant breeding, taxonomy, genetics, and epidemiology, just to list a few. Other disciplines are likewise divided into distinct groups that often consider each other rivals in the pursuit of resources and recognition.
Barriers have also been erected between the various pest science disciplines despite the fact that these groups are, in essence, interdisciplinary. Entomology, plant pathology, nematology, and weed science study the interactions of different groups of organisms and thus require broad backgrounds in the plant sciences, microbiology or invertebrate biology, and environmental sciences. The barriers are best exemplified by the disciplnes' development of parallel nomenclatures that describe similar processes, effectively further insulating the disciplines from each other.
In the agricultural sciences, clustering of research, extension, agribusiness, and agricultural policy around commodities hinders the development of broad solutions to problems. Frequently, funds for problem-solving research are provided by specific commodity groups, and research is restricted to the study of a single crop rather than the cropping system as a whole. Institutional structures, including professional societies and academic departments, and consequently funding patterns for research, are largely responsible for the development of barriers that then become hardened through competition for limited research and extension funds and other types of institutional support and recognition. Events in the past few decades that are unique to the pest sciences are challenging efforts to encourage interdisciplinary cooperation between these groups.
IPM programs are the intellectual antecedents to EBPM. As are EBPM strategies, IPM programs were developed on the basis of reducing pesticide use and integrating all pest-management strategies into a coherent unit. Competition for the limited funds, and the way that these funds were administered, led to rifts between and within the pest-science disciplines. Many programs that were conducted under the rubric of IPM can be criticized for becoming "insect pest management" programs or "integrated pesticide management" programs. It is unfortunate that because of the way IPM programs were implemented, one of IPM's legacies is more, rather than fewer, barriers between the pest-science disciplines.
The goals of EBPM cannot be reached without collaboration across taxon-specific disciplines and across hierarchical levels of organization. Interdisciplinary grants and other incentives for collaboration among groups must be crafted in ways that make researchers accountable for doing truly collaborative work. Interdisciplinary activity in the pest sciences must contend with historically based
divisions at the research and administration levels or the problems that plagued IPM research will be repeated.
The pest science professional societies, such as the Entomological Society of America, American Phytopathological Society, Society of Nematologists, and the Weed Science Society of America, each have members who advocate increased emphasis on biological controls for management of pests; but because these societies have diverse interests, they are not likely to provide the necessary leadership for EBPM. A strong, unified group that encourages multidisciplinary research and provides a forum for discussions of common themes related to EBPM could help to assure future interdisciplinary approaches to biological control. Forums for research and extension interactions between the pest science disciplines can also reverse the evolution toward different nomenclatures describing common phenomena. It is time for a professional forum to enhance communication among all scientists involved in aspects of EBPM.
Coordination of Federal Departments and Agencies
Ecologically based pest management is a topic of interest to a number of federal departments and agencies, and strong federal support is needed if progress is to be made in its implementation. The need for a coalition of credible advocates for EBPM from the research and extension communities is essential; strong leadership at the federal level can help to ensure that efforts in research and extension at the state and local levels are supported and coordinated.
A successful model for coordinating problem-solving research is the National Institutes of Health (NIH). NIH works to provide a national research agenda and funding to support both basic and applied research toward solving specific problems. NIH also acts to focus public attention on the problems being addressed. This type of model has been recommended in the past for agricultural research.
Attempts have been made within USDA to coordinate biological-control activities; this type of effort needs to be encouraged and expanded to include other federal departments and agencies with common interests. Identifying a single group within USDA empowered to speak for the various interests within the department and coordinate their activities would facilitate research and extension efforts in EBPM; this group could also coordinate USDA's activities with those of other government departments and agencies with interests in this area, such as the Environmental Protection Agency, the National Science Foundation, the Department of Energy, the Agency for International Development, and the Department of the Interior.
One issue that will necessitate federal leadership is that of the overlapping
goals—reduction of the use of pesticides—of both EBPM and IPM. The primary differences between EBPM and IPM are (a) the domination of pesticides in IPM and (b) IPM's history of implementation primarily for control of arthropods. When EBPM research proves successful, IPM programs will be the transition between current methods and EBPM. Unfortunately, because of the overlapping goals of the two programs, it may be that, rather than working together, there will be competition between the two approaches for limited resources and recognition. Federal leadership is needed to prevent this problem.
INFRASTRUCTURE FOR RESEARCH
EBPM will not be widely adopted without significant new research funding being made available to address current limitations to discovery, development, and implementation. Significant expansion of research in this area faces obstacles posed by financial and political stresses on local agricultural and natural resource research institutions. Even during the best financial times, institutions are slow to change research directions and methods. The down-sizing that is occurring in many research institutions is constraining the ability of administrators to redirect resources into new programs. This is particularly true when EBPM research competes with widely accepted conventional pest-control strategies, many of which are supported by the pesticide industry. It is cheaper to run trials of conventional pesticides to determine their utility in the local area than it is to identify and optimize application strategies for EBPM.
The need for large local research investments in EBPM research is analogous to the need that led to past investments in animal and plant breeding programs. Initial successes in plant breeding efforts demonstrated how directed breeding could greatly improve crops and domestic animals. Nevertheless, it took, and continues to take, a major investment in research to reap the benefits for agriculture from genetics. Confidence in a lucrative return drove this major investment even though the return often came decades after the initial investment was made. There have been enough successes now in biological-control research to justify a similar research investment in EBPM.
The research infrastructure needed to support a significant EBPM initiative exists both at the state and federal levels. The extensive USDA-ARS and land grant university research systems provide research personnel and facilities that are necessary to develop and implement EBPM. However, EBPM research must compete for resources with other research priorities. Major shifts in resource allocations will be necessary before a significant number of scientists become involved in research leading to EBPM.
Providing adequate research funding to accelerate the development and implementation of EBPM is difficult because of the diversity of research needs. Progress in biological control, for example, has been greatest where the necessary basic information is available or easily obtained. Progress has been most rapid at
the macroscopic level where population ecology and predator-prey models are understood. Progress at the microscopic level has been slow because microbial ecology is still relatively poorly understood.
Research must also extend beyond basic to applied research. The need for applied research is compounded by the site-specificity of many ecologically based approaches. Ecologically based management systems with universal applicability are rare. Site specificity places unusual research, development, and extension burdens on local institutions. EBPM strategies adaptable to local conditions for each arthropod, weed, or pathogen pest are needed. Sharing research responsibilities between states can help reduce costs, but each locality will need to independently identify and optimize procedures for each of the major pests of the region.
The base of ecological information necessary to develop and implement EBPM is much greater than that for conventional chemical pesticides. Widespread implementation of EBPM will come only as a result of political processes through which the public makes it known that alternatives to conventional pesticides must be found.