The control of illicit-drug trafficking and drug use is a difficult and complex process that involves a variety of prevention, control, treatment, and law-enforcement strategies. Eradication strategies for controlling illicit-drug crops are used to target the beginning of the drug-supply chain by preventing or reducing crop yields. Mycoherbicides have been proposed as an eradication tool to supplement the current methods of herbicide spraying, mechanical removal, and manual destruction of illicit-drug crops. Mycoherbicides are developed from plant pathogenic fungi that occur naturally in the environment. Some people regard them as preferable to chemical herbicides for controlling illicit-drug crops because of their purported specificity to only one plant species or a few closely related species. As living microorganisms, they have the potential to provide long-term control if they can persist in the environment and affect later plantings. Research on mycoherbicides against illicit-drug crops has focused on three pathogens: Fusarium oxysporum f.sp. cannabis for cannabis (Cannabis sativa), F. oxysporum f.sp. erythroxyli for coca (Erythroxylum coca and E. novogranatense), and Crivellia papaveracea or Brachycladium papaveris (formerly known as Pleospora papaveracea and Dendryphion penicillatum, respectively) for opium poppy (Papaver somniferum).
In response to a congressional mandate (Public Law 109-469), the White House Office of National Drug Control Policy (ONDCP) requested that the National Research Council form an expert committee to examine the scientific issues associated with the feasibility of developing and implementing naturally occurring strains of the mycoherbicide fungi as a means of eradicating illicit cannabis, coca, and opium poppy crops. The study was also to evaluate the potential human health, ecological, and environmental risks associated with the use of these mycoherbicides and to identify future research and development needed to support their use. The committee was charged with addressing the following issues about the potential use of the proposed mycoherbicides: their effectiveness in eradicating their target plants; the feasibility of their large-scale industrial manufacture and delivery; their potential spread and persistence in the environment; their pathogenicity and toxicity to nontarget organisms, including other plants, fungi, animals, and humans; their potential for mutation and resulting
effects on target plants and nontarget organisms; and research and development needs.
To address its task, the committee reviewed publications and other publicly available information on the three proposed mycoherbicides and relevant publications on related fungi and other mycoherbicides that have been developed against undesirable plant species (weeds). The publications were identified through literature searches and by consulting the ONDCP, the U.S. Department of State, the U.S. Department of Agriculture, and the UN Office on Drugs and Crime. The available studies on the proposed mycoherbicides were few, were not all peer-reviewed, and were primarily greenhouse, growth-chamber, and small field studies conducted under controlled conditions. Those limitations made it difficult for the committee to draw conclusions or to make predictions about the performance of the proposed mycoherbicides in larger field settings and under natural conditions.
On the basis of its review, the committee concluded that the available data are insufficient to determine the effectiveness of the specific fungi proposed as mycoherbicides to combat illicit-drug crops or to determine their potential effects on nontarget plants, microorganisms, animals, humans, or the environment. The questions normally asked before a fungal pathogen is registered as a mycoherbicide in the United States have not been adequately addressed. The committee offers the following assessment of what can and cannot be determined at the present time regarding each of the issues raised in the statement of task.
The degree of control that might be provided by the proposed mycoherbicides, the mechanisms by which they cause disease, and how control of the target plants could be maximized have not been established. Although each of the proposed mycoherbicides has been shown to cause disease in its target plant, disease severity was inconsistent and depended on biotic factors (such as age of plants and strain of fungus) and abiotic factors (such as moisture level, temperature, and ultraviolet radiation). For example, the cannabis mycoherbicide caused plant death in one study but low to moderate disease severity in a second study. For the coca mycoherbicide, published mortality ranged from 35% to 94%, but the background incidence of disease and background mortality in noninoculated plants also were high, sometimes approximating those observed in inoculated plants. Some varieties of cannabis were found to be resistant to the mycoherbicide and exhibited no effects or less severe disease.
In studies of the opium poppy mycoherbicide, a range (6-100%) of leaf necrosis in greenhouse and growth-chamber experiments was reported. Disease severity depended on the type of inoculum used (sexual or asexual spores), the age or growth stage of the plants, and environmental conditions, particularly the dew period and temperature after inoculum application. Plants in early development stages were killed or suffered foliar damage; at more mature stages,
poppy capsule number and size were reduced, and the poppy seeds had lower viability.
The types of diseases observed to be produced by the proposed mycoherbicides in their target plants are wilts in cannabis and coca and blight of the aerial parts of opium poppy. However, the mechanisms underlying the hostpathogen interactions and secondary spread of disease, which are critical determinants of mycoherbicide efficacy, have yet to be documented.
INOCULUM PRODUCTION AND DELIVERY
Large-scale production of the proposed mycoherbicides appears to be feasible, although available fermentation capacity might not suffice to combat illicit-drug crops on a global scale. The production process could be adapted from technology developed to produce microbial biomass for pharmaceutical, food, biotechnology, and other commercial uses. Large-scale production of mycoherbicides for commercial use is typically undertaken by industries that have microbial fermentation capabilities. The process involves the production of large amounts of fungal biomass by liquid fermentation, solid-substrate fermentation, or a combination of the two and has been used to produce commercial quantities of mycoherbicides.
It is difficult to estimate the quantity of mycoherbicide needed to control cannabis, coca, or opium poppy crops. Rough estimates based on the few available studies suggest that tens to hundreds of kilograms of dry formulation per hectare would be required for a single application of the cannabis and coca mycoherbicides. More accurate estimates require tests of the finished mycoherbicides under conditions that simulate field operations. Producing the amount of mycoherbicides required for global control efforts may or may not be feasible in light of cost or technical limitations; no mycoherbicide has ever been produced on such a massive scale. Studies of disease of opium poppy have used liquid spray formulations consisting of spores of the mycoherbicide fungus suspended in water or water amended with a surfactant or a vegetable oil. At the rates reported in publications, hundreds to thousands of liters of liquid-spray formulation (containing billions to trillions of spores) per hectare would be required for a single application of the opium poppy mycoherbicide.
The methods for delivering the proposed mycoherbicides to target sites would affect their performance in the field. The cannabis and coca mycoherbicides are soilborne and root-infecting and would have to be applied to the soil on or near plant roots for greatest efficacy. Several dry formulations (such as pellets) have been developed for this type of application. The opium poppy mycoherbicide, in contrast, would attack primarily aerial parts of the plant, so application of a liquid formulation to foliage would provide the greatest efficacy. For all three mycoherbicides, on-ground application would allow the most precise and uniform application. However, ground applications are unfeasible because of uncooperative and possibly hostile growers, who are likely to try to prevent
application of the mycoherbicides. Aerial application of mycoherbicides from airplanes as dry formulations on cannabis and coca fields could reduce their efficacy because the formulations would be subject to scattering by wind, which would lead to nonuniform, discontinuous placement of the inoculum over the target area and reduce the size of the plant-pathogen interface. A similar limitation applies to aerial application of liquid mycoherbicide formulations on opium poppy. But an even more important limiting factor is the availability of water needed for the liquid formulation and the ability to transport and apply the required quantities to the target area.
Another issue is how long can the mycoherbicide fungi persist in the soil after application. It is important to determine whether the population density of the mycoherbicide fungi would remain high enough and for a long enough period to infect the target crops and whether they could survive in the soil and organic matter at levels necessary to affect later plantings of the crop. Another consideration is whether the mycoherbicide strains would pose any additional risks to nontarget organisms after release, in which case the prolonged persistence of these strains would be a disadvantage rather than an advantage.
Only a few studies on the long-term survival of the proposed mycoherbicides against coca and opium poppy are available, and essentially no data on the proposed cannabis mycoherbicide are available. In the available studies, the coca mycoherbicide strain survived for up to 7 months after application, and the opium poppy mycoherbicide strain survived in treated fields for two growing seasons. Those fungi are indigenous where their host plants are grown and have been linked to periodic, natural epidemics, so at least the fungal strains related to the mycoherbicide strains survive for a long time in the presence of their hosts.
Survival of the mycoherbicide fungi would also depend on the environmental conditions. Moisture (from dew, high relative humidity, or rainfall) for several hours and over several days is usually required with favorable temperatures for the fungi to become established on the target plants. In the case of the specific fungal strains studied (such as F. oxysporum f.sp. erythroxyli), data on moisture, temperature, and other requirements for disease development and survival in soil are based on results with one or a few strains collected from relatively small areas. There is no reason to expect these strains to be adapted to any environment other than the one from which they were recovered; therefore, the strains may not be capable of attacking coca throughout its entire range. But the data are also not sufficient to conclude that strains of the mycoherbicide fungi or the diseases that they cause cannot occur throughout the entire geographic and climatic range where the target drug crops are grown.
The mycoherbicide strains might be able to survive on plants other than their target plants or as saprophytes on decaying organic material. Thus, it is
likely that the mycoherbicide strains would persist at some level as part of the indigenous pathogen population once they are introduced in large numbers into the environment. The pathogens also might be spread from the site of application by wind, water, insect or animal carriers, or infected seeds, soil, or plant material. Conditions that might reduce survival of the mycoherbicides include lack of adequate moisture, extreme temperatures (too high or too low), competition or suppression by soil microorganisms, and other biotic factors.
Once the mycoherbicides are applied, their persistence might be shortened through the intentional application of chemicals, such as fungicides or soil fumigants, by growers. Such a control strategy would be most effective for containing mycoherbicides in small areas but would be impractical or impossible for large areas. The chemicals also could affect indigenous microbial communities. There is some evidence that high population densities of the mycoherbicides could be maintained for several months, but the data do not support the hypothesis that the mycoherbicide strains can persist indefinitely at higher population densities than those of indigenous strains of the same fungi.
EFFECTS ON NONTARGET PLANTS AND ORGANISMS
Because of the complexity of native and agricultural ecosystems, it is difficult to predict or quantify the risks to nontarget plants and organisms accurately. Other fungi, soil organisms, plants, animals, or humans could be exposed to a mycoherbicide strain by several environmental pathways. For example, insects, reptiles, and birds might be attracted to a mycoherbicide formulation as a food source, and wind and rain can carry fungal propagules over long distances. Infested soil, plant material, and seeds can be moved to other crop sites by humans or other vectors. Such dispersal would inadvertently expose native plant species to the mycoherbicides and could pose risks to local ecosystems. The coca mycoherbicide, for example, could cause increased disease epidemics of native relatives of coca plants that could lead to adverse effects on local biodiversity or increase erosion if native coca communities on steep hillsides are reduced in size or density. Inadvertent infection of licit crops of cannabis, coca, and opium poppy could have important cultural and economic consequences.
The proposed mycoherbicide strains can cause disease on their target plants, including those grown legally and those indigenous to the area. However, the few host-range studies conducted with nonrelated species are of little value because they report only that the mycoherbicide strains did not cause disease on particular native plants and crops without providing experimental details (and in some cases even the names of the plants). Furthermore, none of the available studies used a standard, systematic process to select the most relevant plants to test in host-range studies. For example, of about 200 species of Erythroxylum native to South America, only two have been tested for sensitivity to the coca mycoherbicide. Thus, the data are insufficient to conclude that the proposed mycoherbicides would not pose a risk to other plants or crops.
Likewise, no data are available on the effects of the proposed mycoherbicides strains on soil microorganisms, animals, or humans. Such effects could include competition with indigenous microbial populations, diseases resulting from direct infection of animals or humans, and disorders resulting from contact with or consumption of toxins produced by the mycoherbicide strains. Although fungal species and strains related to the proposed mycoherbicides might cause such problems, there are no data to suggest that the proposed mycoherbicide strains would produce similar infections or toxins.
It will be all but impossible to control or contain the mycoherbicide strains after they are released. The strains are living organisms that interact with and adapt to their environment. Their ability to survive, propagate, and disperse beyond the target area depends on environmental factors that can be neither predicted nor controlled. The persistence of indigenous strains of these fungi throughout the native ranges of their hosts is consistent with the conclusion that the mycoherbicide strains are unlikely to be contained or eradicated once they are released.
The committee was asked to consider the potential of the fungi to mutate and the possible consequences, but no data on the mutability of the proposed mycoherbicide strains in particular are available. The potential for these mycoherbicide strains to mutate is expected to be similar to that of fungi in general. The genetic makeup of fungi can change by nucleotide substitution, the gain of genetic material from other fungi, the duplication of genetic material, and the loss of genetic material. Some species of Fusarium are well known for their spontaneous mutations to new morphologies. Gene transfer between distantly related strains of Fusarium occurs under laboratory conditions, and there is circumstantial evidence of such transfers under field conditions between strains that belong to different species. There are no data documenting the occurrence of such events in the Fusarium strains proposed for use as mycoherbicides or any strains of C. papaveracea or B. papaveris. Because indigenous strains of these fungal pathogens are present where the drug crops are grown, it is unlikely that mutations would occur in the introduced strains that have not already occurred in the indigenous population or that the mutations would pose novel risks to nontarget plants or other organisms, including humans or other animals. It is not possible to predict what types of mutations might occur, how a pathogen or target plant might be affected, or whether the mutations would be favored by natural selection.
New genetic variation results from mutation, and mutations can become established in fungal populations by natural selection or by chance. Natural selection results in adaptation to changing environments, including adaptation to new cultivars of a target species or to new host species. Adaptation might occur in fungi that reproduce sexually or asexually. Sexual reproduction allows new
genotypes to be produced more quickly through recombination than through mutation alone, as would occur in asexually reproducing organisms.
The evolutionary processes of mutation and adaptation also apply to the target plants. Plant cultivars resistant to a mycoherbicide could exist or emerge through natural selection and replace their sensitive predecessors. Thus, the development of effective mycoherbicides is a continuous rather than a one-time process.
RESEARCH AND DEVELOPMENT
The data available on the proposed mycoherbicide fungi reviewed by the committee are insufficient to determine the feasibility of their development as mycoherbicides or the risks that they might pose to nontarget plants, animals, humans, microorganisms, or the environment. Additional research is needed to address those concerns. At a minimum, a scientific team with expertise in plant disease epidemiology; plant pathology; fungal genetics; fermentation, formulation, and application technology; and nontarget risk assessment is needed to develop the proposed strains as mycoherbicides and to assess their safety and effectiveness. Initially, research is needed to study several candidate strains of each fungus and to identify strains that are the most efficacious under a broad array of environmental conditions. The resulting information would guide formulation development, the selection of a delivery method, and the scaleup required to generate enough mycoherbicide product to achieve a significant level of control.
Multiple regulatory requirements must be met before a mycoherbicide could be deployed. Many of the regulations focus on evaluating the risk to the environment posed by the introduction of the mycoherbicide. Little research of this sort has been conducted with any of the proposed mycoherbicide strains. Before the mycoherbicides could be used outside the United States, additional regulations in one or more international agreements might also need to be met, including the International Plant Protection Convention, International Standards for Phytosanitary Measures, the Biological Weapons Convention of 1972, and legal requirements in the country where the mycoherbicides are to be used.
Studies of the cannabis, coca, and opium poppy mycoherbicides that have been published or were made available to the panel are preliminary, exploratory, and insufficient to determine their suitability for controlling illicit-drug
crops. The available data do not answer all the questions normally asked before a fungal pathogen is registered as a mycoherbicide in the United States. The rigorous, lengthy testing required by the U.S. Environmental Protection Agency has not yet begun, and conducting the research is not a guarantee that a registered mycoherbicide product will result. Mycoherbicides for the control of illicit-drug crops will face additional difficulties in that the people cultivating the crops will be working to prevent the mycoherbicides from having their intended effects.
International Approval and Cooperation: Mycoherbicides proved to be safe and effective might not be approved for use in other countries. At least some tests of the mycoherbicide strains must be performed in the countries where the mycoherbicides might be used or in other countries that have similar climatic and environmental conditions. The testing requires the approval and cooperation of those countries and has been difficult, or impossible, to obtain. Country-specific requirements for such applications must also be satisfied.
Difficulties in Implementation: Commercial success of mycoherbicides developed to control weeds requires collaboration with the growers. Farmers who welcome attempts to control unwanted plants will tolerate aerial application from aircraft flying at low altitudes and at low speeds or from ground-based equipment, as needed, for the effective application of mycoherbicides, and they will permit or assist in the on-the-ground monitoring needed to assess the efficacy of the mycoherbicide. The proposed mycoherbicides for illicit-drug crops would not have similar cooperation from their growers, and this would constrain aerial application methods and limit on-the-ground monitoring. Technology for the effective application of mycoherbicides from high altitudes has not been developed.
Difficulty in Assessment of Effectiveness: The available data indicate that that proposed mycoherbicide strains are unlikely to kill large numbers of the target plants quickly. The combination of lack of rapid, aggressive action with little or nonexistent on-the-ground assessment would make it difficult, or even impossible, to determine the effectiveness of the mycoherbicide applications.
Development of Countermeasures: Producers of illicit-drug crops have an incentive to prevent damage to their crop yields and should be expected to develop countermeasures that reduce the efficacy of the mycoherbicides. Such countermeasures could include the use of fungicides or soil fumigants to kill the mycoherbicide strains directly or the cultivation of plant varieties that are resistant to the mycoherbicides.
Risks to Legal Crops and Native Plants: Cannabis, coca, and opium poppy are grown in several countries for licit uses and are part of the native flora in some regions. Plants in those settings could be vulnerable to the mycoherbicides. In addition, the mycoherbicides could spread beyond the geographic range of the illicit crops.
Risks to Nontarget Organisms: The mycoherbicide strains could have direct and indirect effects on other plants, microorganisms, animals, or the environment.
Those effects cannot be completely characterized even if research is performed to learn more about the infectivity and toxicity of the strains, if any, to nontarget plants and organisms. Mycoherbicides consist of living organisms that interact with and adapt to their environment, and it is difficult to predict how they might behave when released in substantial numbers into an ecosystem.