Food & Fertilizer Technology Center
5F.14 Wenchow St., Taipei 10616 Taiwan R.O.C.Tel: (886-2) 2362-6239 Fax: (886-2) 2362-0478
Email: info@fftc.agnet.org
Copyright © 1998-2007 FFTC. All rights reserved.
The development of effective long-term control programs for aquatic weeds is dependent upon the ability to integrate biological, chemical, and cultural control strategies. Successful IPM programs are dependent upon a sound knowledge of the cropping system (especially the inputs and outputs of the system), the biology of the aquatic weed pest, and the biology of the control agent(s). There are several ways to analyze these complex systems, from the molecular to the community level. This presentation will emphasize the ecological interactions that should be considered.
Integration of biological control of paddy and aquatic weeds into integrated pest management programs is a necessary goal for the implementation of sustainable rice production systems. To reach this goal, however, a number of obstacles must be overcome. Most important is the current status of biological control of aquatic weeds which, in this discussion, will be defined narrowly as the use of beneficial microorganisms or their gene products for pest control. Although considerable research has been done and several successful systems have been implemented for the control of weeds, these are usually restricted to low-input systems such as public lands and range lands. This is because most of the work has been based on the classical approach to biological control, in which the strategy is to establish pathogens or herbivores in an area where the weed is a problem. Classical biological control is dependent upon long-term population dynamics; however, in intesively managed systems such as Asian rice production, where three crops are grown each year, it is not possible for the long-term interactions necessary for suppression of the weed population to take place. Thus, a more appropriate approach may be to augment or inundate the system with the control organisms at specific times in the cropping history. This is especially true when the control agent is a fungus or bacterium, since these are relatively easy to produce in large quantities and are dependent upon specific environmental conditions for greatest efficacy. When a fungus is used to control a weed in this manner, it is referred to as a bioherbicide (Emge and Templeton 1981). Successful integration of control agents, either pathogens or insects, is dependent upon the compatibility of the biological control system with the entire crop production system.
Most of the difficulties in developing functional biological controls are related to the complexities of the biological systems being managed. In spite of this difficulty, there are increasing economic and social pressures to develop practical biological control strategies. This paper addresses the potential for integration of biological control with other weed control strategies. Also discussed are perspectives of research at population and community levels which should be considered if viable biological control strategies are to be developed.
Biological control for aquatic weeds will be implemented most readily when no other means of control are available, when public awareness of pesticide use is high, and when pesticide applications are particularly hazardous or disruptive to the environment. Such opportunities for implementation are becoming more common. In the past, pesticides appeared to control aquatic weeds relatively cheaply because the negative impacts of pesticides on the environment were costs which were not transferred to the users or manufacturers of pesticides. Recently, a volatile political climate has developed concerning exposure of workers and consumers to pesticides, and there is an increasing body of information describing the negative effects that pesticides can have on cropping systems and surrounding ecosystems. Cosequently, many older pesticides are no longer available for use, and registration of new materials is becoming more difficult and expensive. It appears that despite the complexity of biological control programs for aquatic weeds, such methods will have to be utilized in future management strategies. Fortunately, there are many possibilities for this type of approach.
Integrated pest management has been defined many different ways for different purposes. I will use the definition as proposed in a USDA report (1982) "... (the combining) of two or more pest suppression methods into practical systems of IPM to reduce pest problems." Although other definitions may be more appropriate in other situations, the goal here is to identify opportunities for effectively combining biological control methods with other disease control strategies which are based on chemical, cultural, and genetic methods.
There are many weeds which cannot be controlled effectively with present chemical methods, often because of the toxicity of the chemicals to nontarget species or high costs. Biological control measures might be adopted most rapidly by growers in these instances because no chemical alternative would be available. An example is the use of the rust fungus Puccinia chondrilina for the control of Chondrilla juncea (rush skeleton weed) (Supkoff et al. 1988). Because this weed is important in low input systems, such as range land, it is not economically feasible to control it with chemical herbicides.
Biological methods can also be used to increase the level of disease control attainable with chemicals. It is possible for resident microorganisms to increase directly the efficacy of a chemical treatment. For example, herbicides may stress a plant so that it is more susceptible to infection by plant pathogens, competition from other plants (including the crop plant), or to being eaten by insects. A combination of herbicides with biological control agents may result in the control of weeds which cannot be controlled by chemicals alone, or a reduction in the amount of chemical applied, by reducing either the number of necessary applications or the rate needed for each application.
In each of these instances, the beneficial organisms must have a tolerance to the herbicide applied to control the weed. This is often the case with herbicides, since they usually affect only plants. In some production systems, however, the beneficial organisms may be affected adversely by an herbicide. In such cases, it would be necessary to apply the organisms at suitable intervals after pesticide application, or develop resistant strains of the beneficial organisms.
Biological control measures can also be used to minimize the potential of some herbicides to increase other, nontarget, weeds. In plant pathology, this concept is termed "disease trading" (Kreutzer 1960). For example, applications of herbicides that are specific for broadleaf weeds can lead to rapid infestations of weedy grasses. Herbicide treatment can reduce population densities of these broadleaf weeds which may be important competitors, not only with the crop plants, but also other weed species. A similar situation occurs whenever land is severely disturbed. Many organisms, including weedy plant species, have developed specific life histories to exploit severely disturbed areas. Often these species are considered r-selected species (MacArthur and Wilson 1967), which have characteristic traits such as rapid reproduction, effective dissemination, and tolerance to conditions associated with disturbed systems. These species often do not do well in undisturbed systems, usually because they are poor competitors in communities of high diversity.
Cultural controls are commonly used to create an environment which is not conducive to the development of weeds. Establishment of beneficial organisms within this environment might increase levels of weed control which cannot be attained by the cultural practices alone. For example, water-hyacinth (Eichhornia crassipes) can be controlled by draining waterways for a period of time. However, this method is very labor-intensive and is often not practical, especially when considering the negative impact that such a practice has on the rest of the aquatic environment. However, if the repopulation of the waterway by water-hyacinth could be reduced by beneficial organisms, from fish to fungi, it may be possible to delay the need to repeat such a drastic cultural practice.
Although many situations can be envisioned in which biological control could be integrated with cultural control, the potential for integration may be limited in some instances. In particular, it may be difficult to integrate biological control with cultural practices designed to create environmental conditions unfavorable for a targeted weed, for it may also inhibit the beneficial organisms. For example, irrigation during the fallow portions of the year is an effective way to induce rapid parasitism of weed seeds. However, that very same moisture may also serve as the necessary source of water needed for germination. A similar situation may occur when water is kept from a field to inhibit weed seed germination and growth. Such arid conditions would also inhibit the naturally occurring or introduced pathogens that may otherwise exact a toll on the weed population.
There are instances when it may be useful to integrate biological control measures with plant breeding programs. For example, in some breeding programs it may be possible to select for plants which have allelopathic properties that reduce weed populations.
Within each of these integrated strategies, biological control should assist in management of pesticide resistance within weed populations. Many chemicals have specific modes of action and can rapidly select for portions of the plant population resistant to them. Interactions between a weed and populations of one or more biological control agents should be much less specific, and selection for resistance should be less likely to occur.
In most of the scenarios described, biological control can best be used to augment other weed control strategies rather than to replace them. However, the development of such augmentative biological control methods is difficult because of the complexity of biological systems and the impact of economic and sociological factors.
Biological limitations on the biological control of weeds are directly associated with the very practice of agriculture. As recognized by Feeny (1976), fast-growing plants often rely on their "unapparency" as a defense against natural predators. Unapparency is defined as the period of growth during which the plant is not obvious to potential predators. For example, a plant that is susceptible to attack may minimize the chance of encounter with its pests because it is small or present for a short period of time. In agriculture, this is achieved by planting short-season crops, in which the pests do not have enough time to reach economic levels of infestation. Many of our crop and weed species depend upon a low level of apparency in their natural ecosystems. However, agricultural practices such as monoculture limit this protective factor, since plants become more apparent to pests when they are grown in monoculture systems. At the same time, cultural practices may decrease the apparency of weeds as a result of the lush, high density growth of the crop. By the time the weed biomass is large enough to became "apparent," economic damage to the crop is significant. One would not expect biological control strategies to be effective under such conditions.
Another important aspect of how the biology of agricultural production systems limits the implementation of biological control is the concept of enemy-free space (Price et al. 1980). Organisms, including weedy plants and their associated pests, often disperse or evolve in various ways in order to escape their enemies. In the biological control of weeds, it is imperative that the control agent be introduced and maintained in its own enemy-free space, so that its unrestricted growth will have a better chance of responding to any rapid increase in the biomass of the weed. Goeden and Louda (1976) felt when biological control of a weed fails after introduction of an herbivore, which occurs about 50% of the time, it is due in part to the interference by enemies of the herbivore control agent.
Price (1981) applied the interactions between natality, predation, and intraspecific competition to a three dimensional model. He concluded that for strongly r-selected species (such as weeds) in unstable habitats, natural enemies have little or no effect on the population dynamics of the weed. It is not until more stable, K-selected, communities develop that natural enemies limit a population, which of course does not occur in the intensive management of many crops, including rice.
Much of the research done in the last few decades has not addressed these complexities to the degree required to develop workable biological control strategies. Many organisms have been selected as potential control agents on the basis of limited greenhouse tests. When these organisms are introduced into natural systems with variable environmental conditions, it is very difficult to reproduce the level of pest pressure observed in the intensively managed experiments.
Under these constraints, biological factors determining the classic predator-prey/herbivore-host relationships of population biology which have been applied successfully to the control of insects and weeds in low-input systems are not as helpful when we are investigating the control of aquatic weeds in high-input, rapid turnover systems. Unfortunately, few models exist which describe the complex biological interactions in relation to biological control of aquatic weeds.
The implementation of workable biological control strategies has also been constrained by sociological factors, although more so for the use of beneficial plant pathogens than for beneficial insects. Generally, the complexity of the microbial world and the insidious nature of diseases produced by microorganisms are misunderstood by the public. This is not a trivial situation. It has resulted in extremely broad government regulations that are stricter with regard to the registration and importation of bacteria, fungi, viruses, and nematodes than those for plants and animals. These constraints make research difficult, decrease the alternatives available for selection of possible control agents, and in general, increase the costs of bringing a product to market. A similar situation exists when we try to develop effective biological control agents utilizing the tools of genetic engineering. Only through new government policies and programs will these barriers be eliminated.
The economics of the development of biological controls for aquatic weeds are complex. The classic approach to biological control of weeds is based upon single releases of selected predators or parasites which maintain themselves at population densities high enough to effect economic control. Thus, the major expenses are incurred in the identification and release of the control agent; little or no costs are incurred after release. In contrast, augmentative or inundative releases require that control agents be applied over large areas each time control is required. In this approach costs are incurred throughout the period of control, not just during the research phase as in classical biological control. This type of scenario is often more appropriate for private industry involvement and will require increased cooperation between the public and private sectors.
Solutions to sociological and economic constraints lie largely outside the realm of biology, and will not be discussed further here. However, constraints which are imposed by biological complexity might be better understood if we adopt new approaches to their study. Much has been accomplished at the organism level in developing an understanding of types of interactions which can take place between a plant and its enemies. However, more work needs to be done to develop insight into the types and numbers of interactions which take place at the population and community levels of biological organization. Much more attention should be focused at these levels of examination _ particularly within spatial and temporal frameworks. Because of the limited opportunities to manage weed populatins over extensive areas, many control strategies have focused on the protection of specific fields. This reduces greatly the spatial dimensions of a pathosystem, and is an approach commonly used with chemical control. However, biological control methods may best be developed at the regional or ecosystem level. The use of refuges and staggered crop periods can maintain a high level of beneficial control agents from year to year, thus minimizing the temporal enemy-free space available to the weed.
Development of successful protective strategies will depend on knowledge of the population behaviors of beneficial organisms and the weeds. The in vitro methods of screening for antagonists which have been used historically may not be the best approach. Much more attention must be given to the behavior of potential beneficial organisms in situ. Studies are needed to determine how beneficial organisms survive under adverse environmental conditions, how rapidly they increase their populations, how the maintain these populations under favorable conditions, and how they interact with weed populations under variable conditions.
Interactions between populations of beneficial organisms and weeds do not occur in isolation from populations of other organisms. It is likely that interactions between these groups of organisms in a community will determine the success of a biological control strategy. Therefore, the complexity of an agroecosystem must be considered in biological control research.
An ecological continuum exists in crop production systems ranging from the very simple to the very complex. In a simple agricultural system, such as hydroponic culture, biological diversity is very low and environmental conditions may be relatively homogeneous and strictly controlled. Conversely, in a perennial forest system, there is great diversity in the abiotic and biotic conditions.
Regardless of the complexity of a production system, strategies for optimizing interactions between populations of beneficial organisms and targeted weeds will have to survive severe perturbations of the system. Most agricultural practices severely disturb an agroecosystem, thus interrupting the natural species interactions that would have otherwise taken place. Since classical biological control is more susceptible to perturbations, it is usually assumed that the inundative or augmentative approach would be more appropriate for a cropping system that experiences severe perturbations. However, this may be in part because most biological control agents have extremely specific host ranges. Endemic, broad host range organisms may actually increase their negative impacts on the pest if a perturbation, such as irrigation, were applied at the appropriate time and under the right conditions for the control agent to attack the plant.
It should be evident that the development of successful biological control strategies depends on evaluating systems at many levels, ranging from the overall crop system to the microsite where the interaction occurs. Consideration of behavior at these levels within the framework of population and community concepts should assist in developing principles of biological control which are applicable across a range of agricultural systems. Within this framework it should be possible to move beyond the empirical approaches used so often in the past.
For biological control of aquatic weeds to be a valuable tool in integrated pest management, the approach to the study of integrated pest management itself may need to be redirected. Levins (1986) makes the case for an ecological approach to agriculture. Until now, integrated pest management has been applied to production practices which were developed during a period when pest management in agricultural systems focused on elimination or control of single pests independently. It may be more productive to approach agriculture itself as an ecological system in which diversity, long term dynamics, and social goals are considered within the framework of community concepts.
Not only is the ecological approach desirable, but now for the first time it may be possible. The rapid development of advanced instrumentation, such as field computers which can monitor a myriad of environmental conditions at short time intervals, has had a tremendous effect on how we produce crops and control their pests. The areas of applied and mathematical ecology, combined with the tremendous computing power now available, put time series analysis, geostatistics, simulation models, and complex sampling strategies within the grasp of every grower. These new methods of understanding agricultural communities will enhance the potential to apply biotechnology to crop production. Ultimately, the integration of biological control with other aspects of crop production may enable us to develop production systems which approach the lofty goal of sustainable agriculture.
Dr. Baki asked about the probable reaction of politicians to biological control. Dr. Marois informed the meeting about the influential lobbies in Washington working to promote acceptance of biological methods in USA. He pointed out that government policy in most countries is concerned to minimize the risk of overproduction, as well as the risk of underproduction. Dr. Baki agreed, but felt there is a serious dichotomy between the developed world, where there are powerful pressure groups working to promote research and development in biological control, and less developed countries where there seems to be little or no pressure on governments to promote biological control.
With regard to the protocols for integrated weed management, Dr. Moody pointed out that any system has to be simple. Farmers are unlikely to reject a laborious procedure such as counting weeds in the field. Dr. Marois agreed, and suggested that the most difficult part of research is often simplifying it and explaining it clearly to those who will use the technology. He suggested that advances in computer technology have meant that it is now easy to use computers in highly complex systems. More information specialists are needed, to simplify information into a "sound bite".
Download the PDF. of this document, 38,088 bytes (37.2 KB).