Technical Service Center
Environmental Applications and Research Group — Publications
Biological Control of Purple Loosestrife at Bureau of Reclamation Research Sites
Technical Memorandum No. 8220-99-07
The Bureau of Reclamation (Reclamation) constructs and manages water storage and delivery systems in the 17 Western states. Frequently, the operation of Reclamation water systems and rights-of-way is impeded by dense growth of either aquatic or terrestrial weeds. The Technical Service Center (TSC) provides assistance with the identification and control of these weeds. In keeping with Department of Interior policy (609 DM 1, June 26, 1995), the use of Integrated Pest Management (IPM) techniques is emphasized. These techniques combine the use of chemical, mechanical, cultural, and biological controls.
Purple loosestrife (Lythrum salicaria) is a perennial, non-native plant with attractive purple flower spikes. It outcompetes native species and forms dense monocultures in wetlands due in part to its high reproductive capability (one mature plant can produce several million seeds each year), and the absence of effective native predators. In areas where this noxious weed is found, it rapidly impedes Reclamation water systems and degrades associated wetland habitats.
Three groups of insects are being used for the biological control of purple loosestrife: leaf-feeders, root-feeders, and flower-feeders. Initial Reclamation field research began at three Denver-area sites in 1994: Coot Lake in Boulder, the Federal Correctional Institute (FCI) in Englewood, and Ward Lake Tree Farm in Lakewood. In 1995, work began at a site south of Ephrata, Washington. This site is on Reclamation property at the Columbia Basin Project, Winchester Wasteway and consists of 25,000 acres (1989 aerial survey) of a near-monoculture of purple loosestrife.
This report succeeds Technical Memoranda Nos. 8220-95-1, 8220-96-13, and 8220-97-5. It details the results to date at each of the above sites, and presents information on the mass-rearing of Galerucella and Hylobius insects for redistribution to field sites and other cooperators. Additional journal publications are in progress for results at the Ephrata site, and for the artificial diet used to mass-rear Hylobius.
Galerucella calmariensis and Galerucella pusilla
Galerucella calmariensis and Galerucella pusilla are leaf-feeding beetles in the order Coleoptera, family Chrysomelidae (photos 1 & 2). Both the adults and the larvae feed on the leaves of purple loosestrife. These two species are difficult to tell apart, and will be discussed as a combined group. Galerucella overwinters in soil and plant litter as an adult. Adults emerge in the spring after the loosestrife plants have begun to grow and lay eggs on the leaves and stems. The developing larvae feed on the leaves and apical meristems of the plant, affecting the plants' growth and flowering. The larvae pupate in the soil before emerging as adults the same year in early summer. The number of generations per season varies from one generation in cooler areas to two or possibly more in hotter areas with longer growing seasons, but adults generally live for a year or less.
Initial mass-production of Galerucella in 1994 was done using individually-sleeved plants in the greenhouse, and a starter-colony of Galerucella from Cornell University. Methods evolved over the years to current production in artificial ponds or children's wading pools containing a mass of plants inside a large common tent. Current methods have cut down much of the labor initially needed to mass-produce these insects. The bulk of the labor now consists of digging rootstocks in late autumn, potting plants in April, and harvesting new adults in July.
Roots used for mass-rearing in the 1997 and 1998 seasons were dug at FCI in November, 1996 and overwintered outdoors each winter covered with straw. The roots overwintered well and were potted in the greenhouse the first week in April and moved to an outdoor pond the last week in April each year. The folly of setting up the tents to contain insects too early in the season was shown on April 24, 1997 when a heavy snowstorm left 12" of snow at the Federal Center and caused the collapse of the rearing tents (photo 3). Mass-rearing for 1997 was still quite successful, with over 38,000 Galerucella produced and distributed. Rearing efforts were scaled back in 1998 and produced about 7,200 insects. In addition to releases noted in the following site sections, Galerucella were sent to the Forest Health Protection Service in Boise, Idaho (2,000 insects in 1998).
Insects used to initiate rearing in 1997 had been overwintered on sleeved plants at 40°F in a growth chamber. This was not very successful; only an estimated 5% (150 of 3,000) survived. This number was supplemented with about 100 field-collected beetles which were placed on plants in the outdoor tents on May 1, 1997. Harvesting and redistribution of newly-emerged Galerucella began the first week of July and continued throughout the month. In 1998 about 150 insects used for mass-rearing were collected entirely at local field sites and placed into the tents during the first week in May, with harvesting and redistribution beginning in mid-July.
Galerucella Releases and Results at the Ward Lake Site
Purple loosestrife at this site is concentrated in a ring about 1 meter wide encircling a small lake. Ward Lake is primarily used as a run-off water management basin and is surrounded by private land used as a tree farm nursery. The lake itself is about 230 meters long and 25 to 70 meters wide. Purple loosestrife plants at Ward Lake are typically the first in the Denver area to begin growth, followed by FCI and then Coot Lake. Shoots emerge about the last week of April, and Galerucella adults can be found about ten days later. Egg-laying begins within a few days, in mid-May.
Galerucella adults have been released at Ward Lake as follows: 1994 - 600; 1995 - 326; 1996 - 0; 1997 - 2,200; 1998 - 0. These releases were made into 3' square tents or sleeves placed on individual plants which were removed after one week. Subsequent monitoring was done at a minimum of five permanent 1 m2 quadrats two times per season, timed to give the best data for insect abundance and impact on the plants. Monitoring protocol and forms are in Appendix I. Release sites for Galerucella have been clustered at one side of the lake (north shore), in order to observe movement of the insects around the lake, as well as movement between release sites.
Table 1 shows selected monitoring data from 1995 to the present for Ward Lake. Average numbers of Galerucella found at the permanent quadrats have increased steadily, and are mirrored by a decrease in average stem height and average number of stems per square meter (Graph 1). The insect effects are visible in a comparison of photographs taken of the area in 1994 and 1998 (photos 4 & 5), and other photographs (photos 6,7,8) of the plants and site.
The impact of the insects on the plants increases as the insect population has time to build. The population will then migrate to other areas of purple loosestrife. This movement was easily traced at Ward Lake. In 1996 plants with foliage damage exceeding 50% were found near Quadrat #3, in the center of the north shore of the lake. Approximately 10 m of plants along the shoreline were affected at that time. In 1997, impacts were visible in the same area again, and had spread about 20 m to the east and west along the shore. By the fall of 1998, the heavily affected area had spread as far as 80 m to the west from the initial zone.
Galerucella Releases and Results at the Federal Correctional Institute (FCI) Site
Purple loosestrife at this site forms a discontinuous ring about 2 m in width around a large constructed lake used for water storage and delivery. Use of herbicides to control loosestrife is precluded by downstream uses, FCI policies, and heavy use by local wildlife.
Numbers of Galerucella adults released at FCI have been as follows: 1994 - 1,000; 1995 - 546; 1996 - 0; 1997 - 2,500; 1998 - 0. These releases were made into 12' square tents or sleeves placed on individual plants which were removed after one week. Subsequent monitoring was done at a minimum of five permanent 1 m2 quadrats two times per season, timed to give the best data for insect abundance and impact on the plants. Monitoring protocol and forms are in Appendix I.
Counts and qualitative observations of Galerucella beetles in May 1996 had suggested that the overwintering success at FCI was less than at Coot Lake or Ward Lake. Management practice at the FCI site had been to mow close to the edge of the water and may have left the beetles with only minimal plant litter on dry land to provide cover for overwintering, increasing mortality. Management practices were altered to leave an unmowed border of a few feet for the last several mowings of 1996, and this practice was continued in following years. This may have contributed to the increased number of insects that were seen during counts in June 1997 and 1998.
Table 2 shows selected monitoring data from 1995 to the present for Henry Lake at the FCI. Average numbers of Galerucella found at the permanent quadrats have increased steadily, and are mirrored by a decrease in average stem height and average number of stems per square meter (Graph 2). Significant damage to the plants was initially noticeable over a wider area than at Ward Lake, where the progress of the beetles could be followed around the shoreline. In 1997, plants on the east shore of Henry Lake were showing slight reductions in height and stem number, but were still able to flower. In 1998, there was extremely heavy and widespread plant damage all along the east shore, and no flowers at all. This effect of heavy feeding by Galerucella occurred over 250 meters of shoreline plants in the same year, in contrast to the movement at Ward Lake of 20 then 80 meters per year. The impacts of Galerucella feeding can be seen in photographs taken in 1997 and 1998 (photos 9, 10, 11).
Galerucella Releases and Results at the Coot Lake Site
Coot Lake is the largest local site, consisting of about 15 acres of loosestrife dispersed among other wetland plants in a constructed wetland. This area is managed by Boulder Mountain Parks, which has attempted to control the spread of loosestrife by cutting and burning stalks in the fall to reduce the seed production. Use of herbicides to control this weed is discouraged by local policies. The practice of cutting the stalks in the early autumn causes reduced vigor of the plants and the plants are not as tall as at other sites. Cooler temperatures due to the rural location affect the emergence of the plants and insects, causing them to lag slightly behind the other local sites.
Numbers of Galerucella adults released have been as follows: 1994 - 800; 1995 - 250; 1996 - 0; 1997 - 6,100; 1998 - 0. These releases were made into 12' square tents or sleeves placed on individual plants which were removed after one week. An additional 4,900 were free-released in 1994 by the Colorado Department of Agriculture. Sixty of these latter insects were kept for testing at Reclamation rearing facilities, but no eggs were produced from these. It is unknown if any eggs were produced in the field from these releases.
Two students from the University of Colorado at Boulder were contracted to identify and prepare herbarium specimens of the 35 most common plants at Coot Lake in July, 1997 (Appendix II). Two herbarium sheets of each specimen were made, and one was submitted to Boulder Mountain Parks for their retention. These sheets will be used as a reference for monitoring at the permanent quadrats, and for GIS/GPS mapping.
GIS/GPS mapping was begun at Coot Lake to map the areas and densities of purple loosestrife infestations. Reclamation personnel used a Rockwell Precision Lightweight GPS Receiver Precise Positioning Service to collect the data. These federally issued units are able to internally correct the intentional errors induced to limit the accuracy of other open-market receivers for government security purposes; this makes corrections from a base unit unnecessary. Data was collected May 16 thru June 20, 1996 ending when the height of cattails and loosestrife made further mapping impossible. Susan Miller, Boulder Mountain Parks, initially assisted Reclamation with mission planning and logistics, and also converted the raw data to a map format with photo background. Personnel changes at the Boulder office have delayed production of a final map, but one will be produced this season.
Monitoring was done at five permanent 1 m2 quadrats two times per season, timed to give the best data for insect abundance and impact on the plants. Monitoring protocol and forms are in Appendix I. Table 3 shows selected monitoring data from 1995 to the present for Coot Lake. Average numbers of Galerucella found at the permanent quadrats have increased steadily, and are mirrored by a decrease in average stem height and average number of stems per square meter (Graph 3). Significant effects on plants have not been seen as quickly as at Ward Lake and FCI, possibly because it is a larger infestation of loosestrife with more area for the insects to disperse. However, in 1998 complete devastation of plants in a small area (about 10 m2 ) was seen for the first time at this site. Based on experience at other sites, expansion of this area of affected loosestrife should occur in 1999.
Galerucella Releases and Results at the Winchester Wasteway (Ephrata, WA) Site
The Winchester Wasteway is owned by the Bureau of Reclamation, with the nearest office located at Ephrata, WA. The wasteway functions in irrigation and irrigation return flows. This land was originally desert, but was flooded after the Grand Coulee Dam was built on the Columbia River. Much of the project and surrounding land and water are currently managed by the WA State Department of Fish and Wildlife as wildlife habitat, especially for fish and waterfowl. A 1989 aerial survey of the area estimated about 25,000 acres of a near-monoculture of purple loosestrife.
Several releases of Galerucella were made by state agencies prior to 1995. Numbers of Galerucella adults released from Bureau of Reclamation mass-production programs are as follows: 1995 - 3,252; 1996 - 3,200; 1997 - 27,600; 1998 - 5,200. These releases were made during July/August each year into 3' square tents or sleeves placed on individual plants which were removed after one week. Releases were made throughout the Winchester Wasteway and some specifically at a research site located near Beda Lake on the wasteway and consisting of an isolated, thick, monoculture stand of purple loosestrife entirely on dry land. This stand measures about 90 meters long and 10-15 meters wide.
Subsequent monitoring of insect abundance and plant vigor was made at the Beda Lake site according to monitoring protocols shown in Appendix I. In addition, numerous land-based and aerial photographs of large areas of the Winchester Wasteway were taken in 1995 and 1998 . These photographs (photos 12 - 15) illustrate the phenomenal effects the beetles are having on the loosestrife over a large portion of the wasteway.
Analysis of the aerial photographs for 1995 and 1998 has been completed (Table 4). The data indicate that acreage of loosestrife affected by the beetles is increasing. This is seen in the increases in acreage of dead and non-flowering loosestrife, as well as the decrease in moderate and thick density areas of loosestrife. A journal article is in progress for the results of this analysis. Unfortunately, the aerial photographs also document an increase in the acreage being invaded by another undesirable plant, Phragmites australis (common reed).
In 1998, populations of Galerucella were having such a significant impact on the plants that a decision was made to open the area for collection and redistribution of insects to other areas of the state. As documented in an e-mail sent to Debra Eberts on 7/24/98 by Craig Conley, Conservation Agronomist for Reclamation in Ephrata, this effort was a success:
Last winter I told the Washington State Noxious Weed Board's Bridget Simon that if the Galerucella emerged in 1998 the way we thought they would, it could be a good opportunity for others to collect and introduce to other loosestrife infestations within the state. As you know the Galerucella did come out smoking in April of 1998 and even the overwintered adults actually did a tremendous amount of feeding.
From your experience of rearing Galerucella, I guessed the next generation of adults would be out in good numbers by the week of July 13, 1998. Bridget sent out a flyer to the county weed boards throughout the state saying if anyone was interested in collecting Galerucella we had a site that would tentatively be ready the week of July 13. This led to the Washington State Department of Ag purchasing some coolers, bug cups and other collecting materials to assist in the effort. To get to the site we wanted to use for collection and to show off our major infestation (ground zero) Hugh brought the S&MC/South District boat and Marty Genrich supplied a boat and driver from the State Department of Fish and Wildlife.
We ran a group of people down the Winchester wasteway each of the five days of the week of July 13. Everyone got to see the results of the Galerucella releases and collect insects for their own use. It went well. ... I gave them a brief description of the Columbia Basin Project, the history of the loosestrife infestation at the wasteway and the history of the statewide control efforts with the background of the introduction of the bio releases. After that people drove to the Dodson road access on the Winchester where Hugh and Jeff were waiting with the boats. We would then proceed down the wasteway about 2 miles to ground zero where they could see for themselves the potential of Galerucella as a biocontrol agent. The area is impressive. In 1997 about 5 acres were completely defoliated with absolutely no green showing, even on the stems. As you proceed out from this site (ground zero) you can see the levels of Galerucella effects for a mile in each direction. Thereafter as you proceed down to the area on the wasteway, you start to see some holes in the leaves, some skeletonization, with a steady increase of effects until total control. The area continues to expand and while the larvae are heavily feeding you can actually see a daily progression. Upon reaching ground zero we then went back upstream to kind of the leading edge of the skeletonization and collected adults.
Collections were made using a 2 liter pop bottle with an 8 inch wide plastic funnel duct taped to the top. Shake the plant over the funnel and then using a 2 inch paint brush sweep the Galerucella into the container. We took everyone to the same area every day. Lots of insects.
Participants included the following:
13 Counties (one third of the state)
Asotin, Benton, Franklin, Grant, King, Lewis, Okanogan, Pend Orielle, Stevens, Pierce, Walla Walla, Whatcom, and Yakima
3 Weed Districts
Benton #1, Grant #3, Intercounty #52
WA St. Dept. Natural Resources, WA Dept Fish and Wildlife, US Fish and Wildlife, WA Dept. of Ag, BOR, South Columbia Basin Irrigation District, State Nox. Weed Board, One High School Teacher, Grant County Public Utility District, and WSU.
Plus a lot who want some now or are sorry they missed out. Collecting should be good for a while yet.
Hylobius transversovittatus is a root-feeding weevil in the family Coleoptera, order Curculionidae. The adult weevils feed on the leaves and stems of purple loosestrife, but the larvae feed on the roots of the plant. Adults may live for two or three years, and spend the winters in the soil and leaf litter. The adults are difficult to find in the field because they are cryptically colored and are generally nocturnal. The females lay eggs throughout the summer in the soil around the base of purple loosestrife plants or in a chamber chewed into the stem. Under normal field conditions, these eggs take one to two years to pass through the larval and pupal stages within the root and emerge as adults.
Hylobius transversovittatus production
Initial work with Hylobius began when eggs were received from Cornell University in 1993. Adults from these eggs began emerging and producing more eggs in 1994. A laboratory culture of Hylobius has been maintained at Reclamation since that time. These insects are housed in cages within environmental growth chambers (photo 16). Inside these cages are cuttings of purple loosestrife inserted into water through a piece of floral foam. The adults feed on the leaves of the cuttings and lay eggs in the stems or in the floral foam ("soil") around the base of the cuttings.
Eggs are harvested every seven days by searching for egg-chambers in the stems and by carefully scratching away the floral foam in layers and uncovering the eggs (photo 17). A fraction of the eggs produced in the laboratory each year is placed onto greenhouse plants. After several months these roots are packed in moist sphagnum in specially-designed boxes (photo 18) within environmental growth chambers to await the emergence of new adults. Adults emerging from the rootboxes are not placed into the field, but are kept in the laboratory to continue producing more eggs for placement onto field plants or artificial diet.
Reclamation joined with Cornell University in 1997 to develop an artificial diet for rearing Hylobius. Initial trials demonstrated that the weevils could mature from egg to adult in only 10-12 weeks when using a specific diet and temperature (photos 19 - 22). It normally takes one to two years for Hylobius to mature from egg to adult under field and laboratory conditions when using normal root material and conditions. The feasibility of releasing adults into field sites will be greatly enhanced by using the artificial diet and incubating the eggs in a controlled-environment growth chamber.
Environmental growth chambers at Reclamation's laboratories were used to test two variations of the artificial diet over three temperature regimes in 1997. From these trials, the best diet formulation and temperature were chosen and underwent further refinement at Cornell and Reclamation laboratories during the 1998 season. Data from these trials will be jointly published in a journal article in the near future.
Hylobius transversovittatus release sites
Prior to the development of the artificial diet, all field releases of Hylobius were done by placing individual eggs (obtained from laboratory cultures of Hylobius) either in the soil next to a purple loosestrife plant, or into a cavity created in a stem. This was a time-consuming process that must be completed before the eggs hatch. Eggs hatch ten days after being laid, therefore some may hatch three days post-harvest.
Eggs have been placed at the following field sites as noted:
Ward Lake: 1994 - 0; 1995 - 77; 1996 - 93; 1997 - 0; 1998 - 0
FCI: 1994 - 135; 1995 - 57; 1996 - 107; 1997 - 0; 1998 - 0
Coot Lake: 1994 - 366; 1995 - 200; 1996 - 424; 1997 - 850; 1998 - 0
Additional sites where eggs were shipped included: Tennessee Valley Authority, Center for Biodiversity (Illinois), Great Swamp National Wildlife Refuge (New Jersey), and Forest Health Protection Service (Idaho).
Evaluation of eggs placed into stems at the Coot Lake field site show that only a small portion of them become viable larvae that tunnel as far as the root (this evaluation is done by dissection of the stems very late in the season). In 1997, only 35% of the stems at Coot Lake showed evidence that eggs had hatched and larvae tunneled through the stem as far as the root. In 1996 the rate of success at Coot Lake was only 21%, and 25% in 1995. In 1996 the rates were 68% at FCI, and 64% at Ward Lake. The rate in greenhoused plants is nearly 100% every year. The cause of this discrepancy has not been determined.
Eggs have been in place at field sites long enough for adults to have emerged, but adults have yet to be observed in the field (partly because of their nocturnal habits). However, evidence in the form of a characteristic feeding pattern on the leaves and frass deposits have been found at Coot Lake. We did look for signs of eggs being laid in stems, but this was not found.
A shift away from placement of eggs into the field occurred in 1997, when we began developing the artificial diet media and used the eggs for those experiments. In 1997, adults produced from media were overwintered in the laboratory and taken to the Winchester Wasteway (Ephrata, WA) site in June 1998. The 107 adults released in June produced eggs and larvae that were found in a follow-up visit to the release site in August. An additional 235 adults from artificial diet experiments were sent in several shipments before September 1998. Subsequent new adults emerging after September were overwintered in the laboratory and about 250 of these were shipped to Ephrata in April 1999. The total released to date at the Winchester Wasteway is nearly 600 Hylobius adults.
Several species of Nanophyes, a weevil which feeds on the ovaries and seeds of purple loosestrife, have been approved for release in the United States. Some of these insects were released at FCI in 1996, but survival of these insects after the near-complete elimination of flowering in 1998 is in question. Nanophyes is only available from field collections in New York. Introduction of this insect to Reclamation field sites is on hold until shipments are available, or an opportunity to collect them in the field arises.
Mass-production of Galerucella at the Denver Technical Center will be reduced for the 1999 season, and probably eliminated afterwards. Starter cultures of Galerucella will be shipped to several new cooperators in Idaho.
Field days for collection of Galerucella at the Winchester Wasteway site in Washington are planned for spring and summer. These collections will be organized by Craig Conley.
Production of Hylobius using artificial diet will be expanded in 1999, with most of the adults being shipped to Ephrata, WA. Some of the adults will be released at Coot Lake in Boulder, CO.
Monitoring of Galerucella and impacts of biocontrol on the plant communities at all sites will continue, including the aerial photographs at Winchester Wasteway. Ground-based monitoring at the wasteway will be increased significantly to document this unique situation.
Plans to control Phragmites at the Winchester Wasteway will be discussed. It is extremely important to address the control of the very invasive Phragmites, especially as competition from loosestrife is decreased. Phragmites also forms monocultures, and is even more problematic with regard to choking the waterways than loosestrife.
Revegetation will be an important factor in reducing opportunities for other noxious weed to invade areas where loosestrife has been eliminated. Revegetation plans for selected areas of the Winchester Wasteway will be discussed in 1999. Consideration must be given to the importance of the area to local wildlife, especially fish and waterfowl, and the recreational attractions of the wildlife and area. One initial step will be to survey the area and find out which plants are already present, and the extent of the existing seedbank for desirable plants.
We would like to express our appreciation for the contributions and support from the following:
Bureau of reclamation Research Program WATER Project EE007, "Development of Improved Aquatic Site Pest Management Methods"
Bureau of Reclamation Program Analysis Office
Bureau of Reclamation, Pacific Northwest Region
Dr. Bernd Blossey, Cornell University
Bureau of Reclamation, Ephrata Field Office, Washington
City of Boulder, Colorado, Division of Mountain Parks
U.S. Dept. of Justice, Federal Correctional Institution, Englewood, Colorado
Ward Lake Tree Company, Lakewood, Colorado
Bureau of Reclamation, Remote Sensing Group, D-8260
Credit is given to Joseph Zubricky, a volunteer at the Denver Technical Center (D-8220), who created the photographs of Galerucella and Hylobius insect life stages used in this report (photos 1, 2, 19, 20, 21, and 22).