PART I.  INTRODUCTION

Weeds cause great losses to agriculture and impact many other human controlled spheres, such as waterways, where they limit navigation and recreation.  Weeds are increasingly recognized as invaders of natural environments, where they reduce biodiversity and disrupt and destroy the functions and the integrity of biotic communities.  The management of weeds is one of the most important activities in agricultural and other environments.  Biological control has been recognized as an important tool for managing weeds.  Current or potential biological control agents of weeds are entities such as insects, fungi, bacteria, nematodes, and viruses that negatively affect weedy plants.  The three basic approaches to applied biological control are classical, augmentation, and conservation.  Classical biological control involves the importation and establishment of biological control agents that theoretically reestablish the natural biological control and maintain the pest population at low levels comparable to those in its native range.  Augmentation, which involves activities designed to increase the numbers or effect of natural enemies, can proceed through natural processes once a low concentration of the biological control agent is released, (i.e., inoculative releases), or may involve the mass production and release of unnaturally high concentrations of the biological control agent, (i.e., inundative release).  Conservation involves actions designed to protect and maintain existing populations of natural enemies. 

Biological control techniques should be incorporated into integrated pest/crop management systems.  Classical biological control has a long history of important contributions to the control of exotic weed pests.  However, imported agents often inflict insufficient damage to bring a pest under control in natural conditions.  Augmentation has great potential, provided effective biological control agents can be produced and applied or released in sufficient numbers for effective control at a cost competitive with other potential control techniques.  Conservation can insure that potential pests do not reach damaging population levels, but it is often not aggressive enough to bring a pest, particularly an exotic pest, under control.  The selection of an approach depends largely on the ecology of the plant, the ecosystem (i.e., natural grassland vs. annual row crop), and the desired impact on the plant.

Management of weeds through biological control already has demonstrated its usefulness, and thousands of species yet unidentified may provide scientists and growers with further biological tools to control weed populations.  Increasing pressures from pesticide resistance and the continuing influx of invasive and nonnative plant species require that research on biological control be developed further.  The challenges for biological control researchers are to discover, develop, and deliver agents for a continually changing spectrum of weeds, to continue to develop effective weed management strategies, and to insure that organisms to be released will contribute to the sustainability of our agricultural and natural land.

PART II.  PROBLEMS TO BE ADDRESSED

A.    Agent Discovery and Selection and Risk Assessment

Problem Statement
Rationale.  Research on the biological control of weeds has led to the identification and development of several excellent biological control organisms.  In some cases, biological control is the only feasible weed control strategy.  However, the use of biological control as an option in weed control depends on the availability of suitable organisms, i.e., those safe to use around other plants yet damaging to the target plant(s) and economical.  Consequently, the discovery and selection of additional biological control agents are critical to increasing the use of this sustainable approach to weed control.  Although biological control has an excellent safety record, use of biological control agents is not risk-free; candidate organisms for weed control are plant pests that, unless tested, may damage crop or native plants in the United States.  Many of these beneficial organisms are from other countries or they are native but used in unnaturally high numbers for weed control.  Therefore, it is necessary to determine if candidates for biological control of weeds pose a risk to nontarget, beneficial, or valuable plants.

What is known.  New invasive weeds and most common weeds originate from other countries and therefore often lack sufficient predators or pathogens that can impact their development.  Such weeds in their native environment are often kept in balance by insects or pathogens that have evolved with the plant and reduce its growth and development through predation or disease.  Thus, a weed’s point of origin has proven to be a good source of biological control agents for the plant.  Weeds growing in a geographical area different from their origin can become hosts to insects and plant pathogens endemic to that area.  Such organisms on their own may have limited impact on the plant population as a whole.  However, if applied at high concentrations, these organisms can dramatically reduce the growth and development of the weed.  While insects and plant diseases can be useful tools to control undesirable plants, there are examples of unintentional deleterious effects of organisms that are released into new environments.  Dramatic shifts in the composition of ecosystems can occur when there is a rapid population growth of one species over others in a community.  Consequently, an insect, pathogen, or virus to be used as a classical or augmentative biological control agent must be demonstrated to be benign in all other aspects.

Gaps.  Efforts in entomology and plant pathology during the past half century have advanced our knowledge of insects and plant pathogens that impact crops, but not until the past two decades has there been a major effort to use these organisms to control weeds.  Consequently, information about insects and diseases that damage or infect weeds is usually incomplete or nonexistent.  In many cases, little is known of pests that may exist in countries of origin for weeds now established in the United States.  Therefore, many new organisms or 'natural enemies' of weeds potentially can be discovered and developed as selective biological control agents.  For newly discovered weed pests, information needed to determine the suitability and value of a weed pest is either unknown or limited.  This information includes the systematics (taxonomic identification), biology, life cycle, ease of in vitro production, and overwintering structures.  Beyond the first discovery or initial screening of a potential weed biological control agent, generally little is known of the organism’s specificity or the factors that may cause the organism to express undesirable characteristics and negatively impact nontarget plants or other organisms.  This is particularly true when an organism is released into a significantly different environment; when the efficacy of the biological control agent has been enhanced, (e.g., to extend the organism’s longevity in the environment); or when unnaturally high concentrations of the organisms are released.

Goals

• Identify and characterize new biological agents to control weeds.
• Ensure that insects, pathogens, or viruses used to control established or invasive weeds do not pose a risk to nontarget plants or the environment.

Approach
To discover potential biological control agents for a given weed survey the geographical origin, locations where the plant is a pest, and areas where establishment of the weed has been limited.  Collect, identify, and evaluate insects, pathogens, and other natural enemies of the weed for their potential as biological control agents.  Utilize information gained in studies of the biology and ecology of the target weed to select and evaluate these biological control agents.  Determine if biological control agents under consideration for release damage plant species related to target weeds or to common companion plants of the target weeds.  Study several different scenarios to ensure that the environmental range and density of the biological control agent and target species are included.  Determine if the biological control agent will disrupt the ecosystem by displacing beneficial organisms similar to the organisms being considered for release.  Utilize both well founded protocols and new technologies such as molecular techniques, GIS, and global positioning systems (GPS) to predict the risks of releasing biological control agents.

Outcomes

• A thorough and accurate risk analysis of biological control agents
• The release of organisms that are effective in controlling weeds yet benign to other plants and the environment for a wide range of target weeds

Impact
The availability of safe, effective biological control agents for invasive weeds to ensure the sustainability of natural and agricultural ecosystems

Linkages to Other ARS National Programs
Crop Production (305)
Food Safety (108)
Rangeland, Pasture, and Forages (205)
Water Quality and Management (201)

B.  Efficacy and Mass Production of Augmentative Agents

Problem Statement
Rationale.  Many insects and pathogens are host-specific and capable of damaging targeted weed pests.  Numerous examples of microorganisms (bacteria, fungi, and nematodes) have been evaluated for use as microbial herbicides.  However, only a few have been registered and commercialized.  Major limitations in developing new biological control agents as bioherbicides include insufficient virulence, high production costs, short product shelf-life, and inconsistent weed control under field conditions.  Equally important to the usefulness of an agent is its availability, as augmentation may require large numbers of organisms for release.  Even classical biological control programs may benefit from the availability of large numbers of agents, as it may improve the initial establishment and spread of the agent over a wide area.  Without economical methods to mass produce organisms, many new and promising biological control agents of weeds will not be available, thus limiting pest control options to chemicals and cultural approaches. 

What is known.  Significant advances have been made in knowledge of factors that regulate virulence of fungal and bacterial diseases of crops.  However, much less is known about factors that regulate virulence of plant pathogens that infect weeds.  Formulation is a critical component in the successful commercialization of a bioherbicide product.  For example, two of the most successful microbial herbicides, COLLEGO® used to control Northern jointvetch, and DeVine® used to control stranglevine, are available mainly because of improvements in formulation that overcame storage and stability problems.  Fermentation and rearing methodologies have been developed that allow for the production of very large numbers of fungal spores, bacterial cells, and insects.  Such production scale systems are already being used for production of some weed biological control agents.

Gaps.  Our ability to produce highly effective biological control agents in sufficiently large numbers at a sufficiently low cost to be competitive with other control technologies is a major impediment to the effective use of many of the organisms that have been identified as potential biological control agents.  Our fundamental knowledge in the molecular genetics and biochemistry of virulence in diseases of weeds is lacking; this may explain why many potential microbial herbicides are not developed because of insufficient virulence.  Because of the natural diversity of insects and plant pathogens, formulation, application, and mass production methods successful with one biological control agent often are not applicable or optimal for others.  Because of the fragile nature of many organisms, formulations developed for agrochemicals are not applicable to microbial herbicides.  Consequently, with each newly discovered or selected biological control agent, studies are needed to overcome limitations such as lengthy dew periods required by fungal weed pathogens and the shelf life of an agent, or to optimize media formulations for mass production or delivery systems of the agent to the target weed before a potential agent can become a useful product. 

Goals

• Generate knowledge of factors regulating virulence of weed pathogens.
• Improve the understanding of microorganisms’ response to environmental challenges and develop ways to overcome these through production and formulation.
• Develop methodologies to produce, formulate, stabilize, package, and deliver highly virulent bioherbicide products.
• Improve our understanding of basic rearing and mass propagation requirements for important weed biological control agents.
• Transfer production technology to action agencies or the private sector.

Approach

Utilize biochemical and molecular genetics techniques to identify and enhance virulence traits of potential microbial biological control agents.  Develop methods to optimize nutritional and environmental conditions for the low-cost production of stable, efficacious microbial biological control agents.  A multidisciplinary team approach consisting of plant pathologists, formulation chemists, biochemists, microbiologists, and weed scientists will be used to develop liquid and granular formulations for pre- and postemergence application of microbial herbicides.  Basic in vivo rearing systems will be used to conduct research on the developmental biology and nutritional requirements of insect biological control agents.  This information will then be used to develop artificial media formulations and mass production protocols, and automated/mechanized equipment necessary to produce the organisms in sufficient numbers and at a low enough cost to ensure competitiveness.

Outcomes

• Improved bioherbicide efficacy
• Reduced costs for production and formulation of bioherbicides
• Extended shelf life of bioherbicide products
• Mass production systems for a variety of insect and microbial biological control agents of exotic and native weed pests
• Increased use of biological control to manage weed pests in the United States through transfer of these technologies to action agencies and/or the private sector

Impact 
Increased sustainability of our agricultural lands and natural areas through reduced use of chemicals

Linkages to Other ARS National Programs
Animal Health (103)
Crop Production (305)
Food Animal Production (101)
Integrated Agricultural Systems (207)
Methyl Bromide Alternatives (308)
Plant Diseases (303)
Plant, Microbial and Insect Genetic Resources, Genomics, and Genetic Improvement (301)
Rangeland, Pasture, and Forages (205)
Soil Resource Management (202)
Water Quality and Management (201)

C.    Field Evaluation

Problem Statement
Rationale.  Monitoring weed populations and the impact of biological control agents on those populations is an essential component of a biological control program.  Data from field monitoring are required for a number of reasons.  They provide feedback that can be incorporated into the program to determine if, for instance, further release of the same agent or additional agents might be needed to achieve the necessary level of control.  They also fulfill the scientific obligation to learn from each experience to enhance future projects.  We have a further public obligation to determine the consequences of releasing each new organism into the environment.  Most of all, we must be able to advise users as to the utility of the control measure and to alert them when a new biological control becomes “operational” as opposed to “experimental.”  Designating a specific biological agent as “operational” marks a successful milestone of the program.

What is known.  Impacts of biological control agents occur at two levels:  (1) changes in growth and reproduction of individual plants and (2) changes in the dynamics of the target plant populations.  The effects of herbivores and diseases on individual plants have been investigated in many studies, but little is known about their quantitative impact on host plant populations.  In terms of practical biological control, the matters of interest are the overall impact on the dynamics of the target and nontarget plant populations.  Nonetheless, studies of component effects of biological agents on individual plants and their interaction with other factors also are needed to provide fundamental knowledge about the ecological interplay between plants and their natural enemies.

Gaps.  At the time of their discovery, new biological agents often are unknown entities, and virtually nothing is known about many aspects of their biology.  In the case of insects, this also is true of their behavior.  Also, often lacking is the critical information on the biology of the target plant or the ecology of the ecosystems in which biological control is contemplated.  Therefore, foundational research on the ecology and biology of both the target and the agent is needed to enhance biological control.  Interactions between insects and plant pathogens are not well understood.  Furthermore, it appears that biological agents need not cause severe defoliation or large scale removal of plant biomass to impact and regulate populations.  Agents may play a significant role in regulating populations by modifying the competitive abilities of one plant species over another.  These interactions need to be investigated. 

Goals 

• Provide basis for evaluating success or failure of a biological control program and possible explanations for those outcomes by understanding the biology, life history, and ecology of prospective target species; factors that suppress target species in their native range; interaction between biological agents and plant competition as related to biological control performance; and interaction of insects with plant pathogens and microbes.
• Document long-term changes in population dynamics of target and nontarget species before and after agent release.

Approach
Evaluating the success of a biological control program is a three-phased process.  First, before releasing the agents, document the status of the weed to provide a baseline for later comparison.  Information about the biology, life history, and ecology of the target species needs to be assembled not only in the invaded areas but also in their native habitats.  Second, following the release of the agents, monitor the establishment and spread of the agent population.  A program of redistribution can be established if the organism is slow to disperse to new areas.  Also important is to determine whether additional species are needed to complement the effect of the initially introduced species.  Third, once establishment has been confirmed, compare data on population of the weed and agents to the pre-release baseline data to assess impact of the agents.  This phase involves the development of various protocols to measure impact:  the design of appropriate field experiments, site selection, the selection and measurements of relevant parameters, and analysis of results.  A powerful approach to evaluating the impact of natural enemies often includes the combination of life table analysis and experimental manipulation of host and agent populations.  Impact is measured by comparing the weed before and after agent release, or inside and outside of the release area, although pesticides or exclusion cages also could be used to address fundamental questions of population dynamics and regulation. 

Outcomes

• Improved control of targeted weeds with biological control agents
• More tools for integrated management approaches to weed control
• Enhanced public image
• Smoother regulatory processes

Impact
Increased crop yields, protection of the natural plant communities, and reduced pesticide use

Linkages to Other ARS National Programs.
Crop Production (305)
Integrated Agricultural Systems (207)
Rangeland, Pasture, and Forages (205)
Water Quality and Management (201)

D.  Combining Biological Control Agents

Problem Statement
Rationale.  In a few examples of biological control of weeds, single agents were largely responsible for controlling the target weed successfully.  However, in most cases, multiple agents appear to be necessary to have a significant impact on a weed’s populations.  In fact, this latter scenario may better reflect the reality in native settings of the weed species where multiple insects and diseases naturally consume and infect the plant.  In such cases, the plants may be affected by pathogens and/or insects that simultaneously infect or feed on roots, stems, leaves, flowers, fruit, or seeds.  Perhaps only in aggregate can these influences have an appreciable effect on the dynamics of the plant’s populations.  Furthermore, in most cropping systems several weed species occur together.  In these situations, a single species-specific biological control agent will not be attractive to managers.  A biological control “package” would be necessary to court adoption under such circumstances.

What is known.  Several plant pathogens, which provided excellent control of their target weeds, were never commercially successful because the research required to fit them into a commercial production system was never performed.  Potentially useful plant pathogens have not been pursued because their range of activity has been too narrow to be used in a cropping system.  In classical weed biological control, the potential of insect and pathogen interactions was established in two of the founding efforts in this field: the program against prickly pear cactus and research to identify agents for the biological control of St. John’s wort.  Recent evidence has confirmed the usefulness of this knowledge.  In biological control of weeds with plant pathogens, agents which, when applied alone, were considered to have a marginal effect on target weeds in crops, were not pursued because of a perceived lack of efficacy.  In classical weed biological control, the opposite has generally been true; a general practice developed of importing and releasing large sets of classical biological control agents to exert cumulative “pressure” against a target perennial weed.  This approach has not met expectations.  The dramatic successes seen with classical biological control agents against certain target weeds at specific sites usually can be attributed to one or two agents.  Furthermore, the desired frequency of establishment and level of impact of classical agents against perennial weeds remain to be achieved in numerous cases.

Gaps.  The interactions of crop rotations, tillage practices, organic soil amendments, cover crops, irrigation systems, and pesticide use practices with biological control agents have not been evaluated.  The use of multiple biological control agents in any given system, combinations of plant pathogens and insects, and integration of biological agents with other control measures have not been adequately explored.  In an increasingly restrictive regulatory climate, an improved focus is needed on research that selects for only the most effective classical agents and combinations thereof.  An improved, focused emphasis on pre-release studies for classical agents in programs to control perennial weeds will better utilize a given agent or combinations of agents in concert with other control measures.

Goals

• Incorporate biological control agents, both microbial and insect, into weed management systems that employ multiple tactics for both crop and noncrop lands.
• Improve the understanding of the effects of crop production practices on efficacy and survival of biological control agents to guide decision-making.
• Develop site-specific strategies for application and deployment of agents that are compatible with system requirements.

Approach
A multidisciplinary approach involving field and laboratory research will be used to improve our understanding of the effects of the physical, chemical, and biological interactions between biological control agents and between agents and other weed management tactics.  Multiple tactic weed management systems can then be developed that address site-specific weed control needs while incorporating biological control agents where they are most useful and effective.

Outcomes 

• Fundamental scientific knowledge concerning the interactions of insects, insects and plant pathogens, and multiple plant pathogens deployed for biological control of weeds
• Enhanced weed control using information on the interactions of insects, insects and plant pathogens, and multiple plant pathogens to develop multiple biological control agent approaches
• Use of biological control agents in systems where they have not been incorporated previously
• Weed management programs that increase the use of biological control agents and nonchemical weed control approaches

Impact
Effective weed control using multiple biological control agents with compatible chemical, cultural, and physical weed control measures

Linkages to Other ARS National Programs
Integrated Agricultural Systems (207)
Methyl Bromide Alternatives (308)
Rangeland, Pasture, and Forages (205)