Around the turn of the century, a European plant that is somewhat toxic to cattle and sheep was first reported in northern California near the Klamath River. It increased and spread rapidly, so that by 1944 "Klamath weed" was found on over two million acres of range land in thirty counties. Land values decreased so much because of the presence of this weed that ranchers couldn't get bank loans for improvements. Herbicides to control the weed were costly and much of the infested land was inaccessible for spraying anyway. The idea of controlling Klamath weed by importing insects that fed on it was proposed as early as 1922, but authorization was not obtained until 1945. By then it was impossible to consider importations from Europe because of World War II. Fortunately, Australian scientists had been working on biological control of the same weed problem in Australia since 1929, so it was easy to import three species of beetles (originally from England and Europe) that had shown promise for controlling Klamath weed in Australia. One of the three beetle species did not survive, but the other two made it through all the tests in quarantine. One was released in the spring of 1945 and the other in February 1946. Both quickly became established; their offspring were then distributed throughout infested areas. More than 3 million beetles were collected for redistribution in California in 1950 from one colony where 5000 beetles were released in 1945-46! Beetles were also successfully established in Washington, Oregon, Idaho and Montana. Within ten years after the first release, Klamath weed was just an uncommon roadside weed in California. Land values increased, and cattle and sheep were no longer in danger. It was estimated that the beetles saved the California agriculture industry about $3.5 million dollars per year in 1953-59, and that these saving continue to accrue each year.

The story of Klamath weed is just one example of successful classical biological weed control. Many of the other success stories of biological weed control until about 1960 were also on alien perennial species infesting grazing lands, but more recently annual and biennial plants, aquatic and semi-aquatic weeds, and native weeds have been the target of biological control research, too. Biological control of weeds historically has usually been considered only after all other management options have been unsuccessful, often decades after the weed is discovered. By that time, the weed can cover immense areas with dense, intractable populations, and there can be thousands of seeds per square foot. Globally thousands of plant species have been accidentally or deliberately introduced into areas outside their natural distribution without their co-evolved natural enemies. Despite sophisticated attempts to exclude potentially weedy species, new species are introduced each year. Some of these introduced plant species will establish and become weeds of agriculture, forestry, pasture, urban situations, or natural areas. With increasing restriction on use of herbicides, classical biological control is often the only choice for managing these weeds in a sustainable, environmentally-compatible manner. Biological control is not a panacea, but is one of the options that should be considered when faced with a new weed problem. When exclusion, eradication, or a quick kill of a weed population is the goal, classical biological control is generally not applicable because the host must be present, classical biological control is relatively slow-acting, and results cannot be guaranteed.

The terminology of biological control of weeds is similar to that used in biological control of other pests, but the terms are sometimes used in slightly different ways (see table below). Classical, or inoculative, biological weed control was recognized and practiced first, with the first modern use occurring in the late 1800's. It remains the most-often practiced type of biological weed control, and is generally conducted by government agencies over large areas. Classical biological weed control relies on establishment by "inoculating" relatively small numbers of individual agents at release sites, from which agents spread on their own. Classical biological weed control is thus essentially an ecological approach to pest management. Arthropods, nematodes, vertebrates, and increasingly, microbial species, are used as weed biological control agents. It is important to use the most appropriate type of agent to match the needs of the program. Classical biological control is limited to importation and release of host-specific natural enemies, and often works best in relatively stable systems, such as pastures. The major risks are known and evaluated before agents are released in the environment. If the target weed is closely related to a crop species, the number of usable natural enemies will be greatly reduced. Successful programs result in cost-efficient, sustainable management of the target species with no or minimal non-target effects.

Strategies of Biological Weed Control
Type	Description	Example
Classical or Inoculative	Introduced natural enemies against introduced pests. An ecological approach.	Chondrilla rust (a fungus) against skeleton weed.
Augmentative or Inundative	Introduced or native natural enemies against introduced or native pests. A technological approach.	Mycoherbicides against crop weeds.
Conservation of Natural Enemies	Enhancing or protecting naturally occurring natural enemies. An ecological approach.	Eliminating pesticides that interfere with weed biological control agents.
Broad Spectrum or Grazing Management	Polyphagous natural enemies confined to ensure restricted action.A technological approach.	Confining goats on leafy spurge or blackberry, or fencing grass carp in canals or ponds.

Augmentative, or inundative, biological weed control was recognized as a possibility about the same time as classical biological control, but was not used widely until much later. Technology is developed for producing relatively large numbers of agents and "inundating" populations of the target weed in the field at a time when it is most susceptible to attack. Commercial formulations of microbial agents used in this way are known as "mycoherbicides", which is an extremely promising method. Most augmentative biological control of weed agents do not become established after release, so must be reapplied each season. Thus, augmentative release is essentially technological in nature. Conservation and broad spectrum biological weed control are more recent developments. They are essentially ecological and technological approaches, respectively.

Although all types of biological weed control should be considered when developing an integrated management program for a particular weed, only classical biological weed control will be discussed further in this article.

Once a particular weed is targeted, there are several steps that are generally followed in a classical biological weed control program. During the "pre-release" or "pre-introduction" (of the biological control agents) stage, extensive information is gathered on the weed and all its known natural enemies. This information is submitted to appropriate State and Federal agencies to obtain approval for the project, and then foreign exploration for good natural enemies can begin. Candidates for introduction as control agents are studied extensively, along with the ecology of the target weed. Promising agents are tested for host-specificity, generally under artificial conditions in laboratories, greenhouses, cages, or field plots in the weed's native area. Once these exhaustive studies have been completed, the data is submitted to the USDA's Animal and Plant Health Inspection Service (APHIS) and the Environmental Protection Agency, who share responsibility for approving weed biological control agents. The last step in the pre-release stage is to import the approved natural enemies to a quarantine facility where they are reared through one or more generations to be sure they are not diseased or have any of their own parasites. In the "post-release" or "post-introduction" stages, the natural enemies are massed-reared and released in the field. Developing the mass-rearing techniques is often one of the most difficult parts of the whole operation, and sometimes the agents can only be distributed in very small numbers at first. (This is also why classical biological control researchers often refer to the method as "inoculative" because even with the best rearing methods, often there are only a few hundred individuals available for release.) Once the natural enemies are in the field, they should be monitored to determine if they become established and if they are having any impact on the target weed population. Even though this is an important step in the program, it is often neglected because funds are frequently available only for discovery and release of agents, but not for evaluation and monitoring. Finally, each natural enemy species is redistributed to other areas in the target weed's distribution, if spontaneous self-dissemination is not adequate.

There have been several recent policy developments that should result in increased support for biological weed control programs in the coming years. In 1990 the National Biological Control Institute was established as part of USDA's APHIS to provided leadership in the area of biological control. In 1992 USDA's APHIS announced a "Biological Control Philosophy" that stated support for biological control as the base strategy of integrated pest management and indicated the agency would "develop regulations that facilitate the release of safe biological control agents." This policy requires fundamental changes that are now being implemented. The USDA Forest Service (FS) and Department of Interior announced major policy changes to "ecosystem management", and in 1993 USDA FS established the National Center for Forest Health Management in Morgantown, WV. The Center has established biological control as one of the key strategies in their program development. In June 1993 the Clinton Administration announced a major pest management policy change: the US will reduce pesticide use by increasing biological and cultural controls. In addition, several major reports have been released recently highlighting the importance of a shift to biological controls from pesticide use, and a "biological control philosophy" has been espoused by many organizations world-wide.

Biological weed control has been practiced for over 100 years, but is underutilized in the United States. Many weeds of crops, pastures, rangeland, aquatic and natural areas have been managed with relatively host-specific biological control agents. With the strong new global philosophical basis for biological control, perhaps many more weeds will be controlled biologically.

- Ernest S. Delfosse, Director, National Biological Control Institute