MARINE PROGRAMS
Strategic Priorities

Biological Control of Weeds Projects

What is Biological Control?

Safety in Biological Control

Biological Control of Spartina

Biological Control of Knotweeds

References

Biocontrol Staff

"Invasive Spartina in Willapa Bay" website

"Olympic Region Harmful Algal Blooms" website

Developing a Control Program for Invasive Knotweeds

The Olympic Natural Resources Center has taken part in the development of a biological control program for knotweeds since 2003.  Current partners in this program include the U.S. Forest Service’s Forest Health Technology Enterprise Team, Cornell University, CABI-Biosciences-UK, and others.  A prior initiated biological control program for knotweeds in Europe has contributed extensive research and exploration that will be useful for a North American program.

The development of a weed biological control program is a 6 step process:

Step 1: Identifying the problem
Step 2: Surveys of existing natural enemies
Step 3: Exploration for candidate agents from the native range
Step 4: Host specificity testing
Step 5: Technical Review and Permitting
Step 6: Implementation

1. Identifying the problem

Japanese knotweed (Fallopia japonica) is listed among the “world’s worst invasive species” by the World Conservation Union. This plant, along with a close relative, Giant knotweed (F. sachalinensis), were introduced into North America as garden ornamentals in the 1800’s.  They have since naturalized throughout the temperate zone.  A fertile hybrid between the two, called Fallopia x bohemica, is even more abundant and invasive than the parent species (Zika and Jacobson 2003).  Knotweeds are herbaceous perennials forming dense thickets of tall stalks 2 to 4 m high with large leaves.  All three invasive knotweeds have spread most aggressively along rivers and streams, but they are capable of occupying an alarming diversity of habitats.  Listed as B noxious weeds in Oregon and Washington, they are slated for control, but are generally considered too widespread for complete eradication.

Knotweeds crowd out native plants through shading, nutrient competition, and allelopathy.  The dense stands have no known value for wildlife, harbor few invertebrates, and can prevent trees from growing near the streams.  Knotweeds can cause increased erosion (Child et al. 1992) as well as increased flooding (Welsh Development Agency 1991) and their forceful roots and rhizomes can cause damage to road surfaces and building foundations (Shaw and Seiger 2002).  Finally, knotweeds are a recreational nuisance, limiting stream access for uses such as fishing and boating.

2. Surveys of existing natural enemies

Prior to importing exotic biocontrol agents, it is important to determine whether any existing natural enemies are already contributing to control and whether any natural enemies from the target weed’s native range are already present as accidental introductions. 

In 2003 and 2004, we carried out surveys of existing natural enemies of knotweed in New York State, Oregon, and Washington.  More than 50 different species of herbivorous insects were found.  Very few of these were found repeatedly or were clearly using knotweed as a host for feeding and reproduction. 

Notable East coast natural enemies include Japanese beetle (Popillio japonica) and an unidentified pathogen.  On the West Coast, notable herbivores include spittle bugs (Philaenus spumarius), woolly bear caterpillars (Isia isabella), a leaf beetle (Galerucella nymphaeae), 3 aphid species, the blue-green sharpshooter (Graphocephala atropunctata), slugs and snails.

All of these are generalist species.  None of the candidate agents from Asia were found. No root or stem feeders were found.  In all cases, damage levels were low.

3. Exploration for candidate biocontrol agents from Asia

Exploration for potential biocontrol agents was carried out in Asia by CABI-Biosciences (U.K.) (http://www.cabi-bioscience.org/) and Japanese collaborators.  A total of 175 herbivores and pathogens were found on Japanese knotweed.  Of these, 7 or 8 were selected for further study as potential biocontrol agents.

4. Host specificity testing

To ensure that the candidate biocontrol agents for knotweed will be safe to introduce, they will be tested for their ability to feed and develop on native and economically important plants in North America with an emphasis on plants closely related to Fallopia spp. (family Polygonaceae).  The work will be carried out in a USDA-APHIS-certified quarantine facility located at Oregon State University.

Prior host specificity testing by the European biocontrol program has simplified the process for the North American program.  Their results have already narrowed down the list of candidate agents to 2 or 3 that appear host specific.  Moreover, many of the plants they tested also occur in North America.  The U.S. program will need to test fewer plant species than if the program was started from scratch.

Our test list currently includes ~60 plant species in 7 categories recommended by the Technical Advisory Group on Biocontrol of Weeds (TAG). 

Categories of plants to be tested
1) Varieties of the target plant
2) All plants in the same genus as the target
3) Economically or environmentally important plants in the same family (Polygonaceae)
4) Threatened or endangered species (or surrogates)
5) Species in the same order (Polygonales) with biochemical, morphological, or phylogenetic similarities
6) Species in other orders with biochemical, morphological, or phylogenetic similarities.
7) Species used by close relatives of the candidate agents

5. Technical review and permitting

Since 1987, weed biocontrol programs in the United States require a rigorous review by the Technical Advisory Group on Biological Control of Weeds (TAG).  This panel is composed of members from 15 environmental agencies from the U.S. plus representatives from Mexico and Canada. The TAG reviews a “petition to import” prepared by biocontrol specialists which includes detailed taxonomies of the agent and target plant, a carefully compiled plant test list, results of host range testing, and discussion of any potential non-target impacts.  Past and current TAG petitions along with TAG recommendations and APHIS action are listed online: http://www.aphis.usda.gov/ppq/permits/tag/.  APHIS-Plant Protection and Quarantine then uses the TAG review and input from individual states in deciding whether to issue the permit.  The TAG review has provided an effective way to ensure that only safe biocontrol agents are introduced.

6. Implementation

Initial Establishment of Agents
Releases will be made first into a limited number of locations in both the Northeast and the Northwestern states.  Initial establishment can sometimes be difficult and so it is important to carefully plan the location and timing of releases.  Mass rearing of the agents prior to release can be valuable, particularly where large releases are needed to avoid Allee effects.  Using a larger number of initial release sites can help eliminate risk of failure due to environmental variability.

Population Growth and Redistribution
The initial release sites will serve as “nursery” sites further collection and redistribution once the populations build up.  It typically takes several years for the agents to become abundant at the initial release sites.  Although the agent will spread on its own, human aided transport will greatly expedite the process. 

Impacts on the Weed
Successful weed biocontrol programs begin with localized damage to individual plants and finish with a sustained regional reduction in the plant population.  The impacts of biocontrol on the weed population can sometimes be dramatic (>99% reduction), but in other cases the level of control is subtle, or variable from site to site.  The likelihood of overall success increases with the number of different agent species used.  It can take up to a decade or more for the full impact of the biocontrol program to be realized.  Quantitative measures of plant population changes over time are crucial to demonstrating the beneficial effects of biological control.