Bureau of Reclamation Banner

Technical Service Center
Environmental Applications and Research Group — Publications

Biocontrol of Saltcedar: Background and Considerations for Possible Use on the Canadian River, Texas

Technical Memorandum No. 8220-01-1
Debra Eberts

October 16, 2000

General Concepts of Biocontrol
History and Present Status of Saltcedar Biocontrol Efforts
Saltcedar Leaf Beetle, Diorhabda elongata
Phases for Implementation of Biocontrol at the Canadian River
Possible Advantages and Disadvantages of Using Biocontrol for Saltcedar


This report discusses the potential biocontrol of saltcedar using the imported insect Diorhabda elongata, and the possibilities for implementation of a basin-wide biocontrol program on the Canadian River, Texas.

Saltcedar is a highly invasive shrub that has invaded thousands of acres in the Canadian River Basin of North Texas. It has been strongly implicated there for causing repressed water supplies - affecting downstream and groundwater availability for irrigation and municipal water. It is also suspected to have severely reduced habitat value of the riparian corridor for many native fish and wildlife species.

Traditional methods employing mechanical removal and herbicide applications for controlling these large stands of shrubs could be costly both environmentally and economically. Costs in 1994 (Sisneros, 1994) ranged from $100/acre for aerial application of herbicides to $771/acre for labor-intensive application of herbicides to cut stumps. Ground-based foliar application of herbicides with backpack sprayers or ground rigs was intermediate in cost ($150-$300/acre). Follow-up costs vary and include herbicide retreatments and any necessary revegetation. Herbicide treatment of resprouts must continue for at least five years, and reinvasion from upstream seed sources is a continual possibility.

There are also a number of environmental costs associated with the various traditional methods of saltcedar management. Aerial application of herbicides will likely kill nontarget, natural vegetation over a wide area. Overspray from ground rigs can also kill desirable plants. Mechanical control by chaining and bulldozing is also nonselective, removing all vegetation and causing widespread soil disturbance. Absence of vegetative habitats adversely impacts wildlife, and renders the land vulnerable to invasions by other noxious weeds. Runoff of herbicides or erosion of unprotected soils can affect water quality in the associated watershed. Additionally, this area is a mosaic of private and public land, making an effectively coordinated saltcedar management action both complicated and difficult. Clearly, to attempt recovery of the Canadian River Basin by these traditional means would be a very costly, laborious and prolonged ordeal.

Ongoing studies of saltcedar biocontrol using insects are demonstrating that the insects have great potential to be more economical and environmentally compatible than conventional mechanical and herbicide control programs. A basin-wide saltcedar biocontrol approach offers satisfactory alternatives to many of the above disadvantages. Biocontrol insects are host specific for saltcedar; that is they consume only the target species. Consequently, soils, other plants, insects, animals, and people would not be directly, or adversely affected by their use.

Costs for starter colonies of insects (Diorhabda and other species as they become available) are substantially more economical since the insect populations can be strategically started and allowed to expand on their own. Insects would be free to seek out saltcedar food sources freely, independent of jurisdictional boundaries. "Application costs" similar to herbicide application are minuscule, and unlike herbicides, the insects should become established and provide long-term control. Monitoring of the insects and their effects, subsequent followup releases, and possibly revegetation, account for the greatest expenditures in a biological control program.

General Concepts of Biocontrol

Biocontrol is man's use of a specially chosen living organism to control a particular pest. New species are constantly arriving in the United States, accidentally or intentionally. When a new species is brought to the United States, its natural enemies are usually left behind. If an introduced species becomes a problem, importing some of its natural enemies can be a way to restore balance and reduce the amount of harm the pest causes. "Classical biocontrol" involves traveling to the country or area from which an introduced pest originated and returning with some of the natural enemies that attacked it and kept it from being a pest there.

Just as the pest and control species existed in their native country in a balance between "predator" and "prey," we expect that balance to redevelop when they are reunited in the United States. Thus, we do not expect a biocontrol agent to completely eliminate its prey, but only to reduce the population of the pest to a tolerable level. Biocontrol is not a quick solution to a problem. It may take several years for the control agent to build up a large enough population to have a measurable effect on the pest species.

The process of importing biocontrol agents is tightly regulated by the United States Department of Agriculture (USDA). Other federal agencies also have a voice in the process through the Technical Advisory Group for Biocontrol Agents of Weeds (TAG) which comments to the USDA on petitions to import and release biocontrols.

Scientists wanting to use classical biocontrol against a pest first determine the origin of the introduced pest and then search for appropriate natural enemies associated with the pest there. These natural enemies are tested to determine their host-specificity. A host-specific agent will not eat or attack any species other than the target species, even if it will die otherwise. This ensures we will not import an agent that will attack any agronomic, native, or desired species once its intended target is controlled. Biocontrol agents that pass the host-specificity test undergo a rigorous quarantine process (to guarantee that no unwanted diseases or parasites are introduced). Only then are they produced in large numbers and released. Follow-up studies are conducted to determine if the natural enemy successfully established at the site of release, and to assess the long-term effects of its presence.

History and Present Status of Saltcedar Biocontrol Efforts

In 1987, the USDA-ARS at Temple, TX initiated a program to investigate biocontrol agents for saltcedar using insects from the plant's native range. In 1994, petitions for field releases of two insects were submitted to the USDA. The technical advisory group to the USDA recommended that both insects be released. Final approval was underway when, in 1995, the southwestern subspecies of the willow flycatcher (Empidonax traillii extimus) was placed on the endangered species list and found to be nesting in saltcedar in parts of AZ, NM, and NV. These flycatcher populations were so low that the USFWS wanted to proceed very carefully with a biocontrol program, even though saltcedar's displacement of cottonwoods and willows is a likely major cause in the decline of flycatcher populations. USFWS decided to allow beetle releases only at sites far from where the flycatcher was nesting in saltcedar and to require careful research on the capacity of the beetles to damage saltcedar and disperse.

The final agreement between USDA-ARS and USFWS calls for research to be conducted at eight sites in six states (CA, CO, NV, TX, UT, and WY). Research will be performed by USDA-ARS and cooperators from universities and other federal and state agencies. These researchers and other interested parties have formed a group known as the Saltcedar Biocontrol Consortium. This Research Phase will be divided into two stages. Stage A research will be done for at least one year inside secure field cages at the sites and will be designed to determine reproduction, impact of feeding on saltcedar, and insect establishment in different climatic zones. During Stage B research, insects will be released from the cages at the same eight research sites. Studies will continue for at least another two years to determine the rate of insect population increase, mortality factors, degree of control of saltcedar, effects on non-target plants, and rate of insect dispersal. The recovery of native vegetation and wildlife after control will be monitored at the original sites for at least ten years.

A decision on the future use of the biocontrol insects at other sites (Implementation Phase) will be made after evaluation of data from the (approximately) three year Research Phase. This decision may allow free and open use of the beetles at any site in the United States, or it may place limits on the use of the beetles. These limits may be geographic and include requirements for monitoring and revegetation.

Outdoor, secure "nursery cages" were approved at Temple, TX and Pueblo, CO in 1997. The plan was to increase the available populations of the biocontrol insect and have sufficient numbers on hand when approval for field cages at all sites was received. Approval to begin research at all the selected sites was granted by federal and state agencies in July 1999. One year of Stage A research has been completed at all sites, and more than that at several sites. Results from all sites are satisfactory, showing that insects can survive the winters and reproduce, and can build to population levels that defoliate trees inside cages. Baseline information has been collected at the Stage B sites about target and non-target vegetation, and on wildlife use of the sites. Researchers anticipate approval to begin Stage B research outside the cages in Spring 2001.

Saltcedar Leaf Beetle, Diorhabda elongata

Diorhabda elongata, a beetle imported to control saltcedar, was originally found in western China and eastern Kazakhstan. In China, where saltcedar is planted to stabilize sandy desert soils, the beetle is common and must be treated with insecticides to prevent it from defoliating and killing the plantations. The larvae and adults damage saltcedar foliage by removing and eating sections of the leaves and also by scraping tissue off the leaves and green stems. This scraping causes large sections of the stems beyond the damage to dry up and fall off. D. elongata can, therefore, cause the death of more plant tissue than it actually consumes.

Both adults and larvae (the young caterpillar-like stage) of D. elongata feed on the foliage of saltcedar, their only known host plant. The adult is 5-6 mm long and is bright yellow with two black strips on its back. Females lay their eggs singly or in clusters of up to 14 on tender leaves. In about 7-10 days the eggs hatch into larvae. The youngest larvae are black but the older, larger larvae have a broad yellow stripe on each side. The larvae molt twice as they grow and reach about 8 mm in length. When finished feeding, larvae drop to the ground and pupate (change into adults) in litter or loose soil.

Results of Stage A research in six states indicate that there is some variation in the length of time their life cycle can take. The time needed for development from egg to adult ranged from 26 to 49 days, probably due to different temperatures at the sites. At sites (or in years) with hotter temperatures, development is completed more quickly than at cooler sites.

Diorhabda elongata overwinters as an adult beetle under leaf litter and in cracks in the soil. There are also site differences in when the adults emerge from and begin overwintering. In China, adults become active in mid-April and begin overwintering by the end of September. This is similar to some of the warmer research sites. At the cooler sites, where foliage emerges on the trees later, the beetles may not emerge until the beginning of May. There are usually three generations per year in China, but we have only seen two per year so far at most of the sites.

The net reproductive rate inside cages (Stage A) probably will differ from that found outside the cages (Stage B). The cages may protect beetles from predators, but it might force adults to lay more eggs than normal per tree, and it also prevents them from migrating to new trees if food becomes scarce. Large numbers of both larvae and adults have starved and died inside cages when this has happened. Taking these mortality factors into account, we determined that we had at least a 25-fold increase in numbers in one generation at the Pueblo, CO cage site.

The actual level of control in the United States cannot be predicted, although the beetles have significant impacts on saltcedar in their home range. Trees in the cage studies have been defoliated several times, but most resprouted. Likewise, the rate of spread of these flying beetles cannot be predicted accurately at this time. More data will be collected in Stage B to help answer these questions, and will be reviewed by USDA and USFWS before release to non-experimental sites is allowed.

Phases for Implementation of Biocontrol at the Canadian River

Implementation of saltcedar biocontrol at the Canadian River can be broken into six stages (although timing of some stages may overlap): determination of suitability, gathering baseline data, establishment of an insect site, redistribution of insects, continued monitoring and data collection, and rehabilitation of impacted sites.

Phase I. Determination of Suitability - Politically and Biologically
Stage B research is expected to end Fall, 2002. At that time the USDA and USFWS will examine the data and decide the future of the program. They may request more testing at the established research sites, or they may allow the release of insects anywhere in any state. It is more likely that releases will be allowed, but with certain geographic restrictions based on proximity to southwestern willow flycatcher nesting sites. The Canadian River site does not lie on any drainages to these sensitive areas, but in direct mileage it may be closer than will be allowed by federal decisions for the initial Implementation Phase releases. So, the "political" suitability of the Canadian River site can't be determined at this time.

Determination of the "biological" suitability of the site will require a site visit during a typical growing season. Two main considerations are the amount of litter or loose soil available to pupating larvae and overwintering adults, and the frequency of flooding. The insects will not be able to establish where floods or permanent above-ground water do not permit insect pupation or overwintering. Biocontrol insects may still be able to affect these areas, but only by adults flying in from suitable habitat to lay eggs on foliage. No adults would develop from these eggs if there is a lack of unflooded soil where larvae can pupate.

Phase II. Gathering Baseline Biological Data
Baseline biological data should be collected the year before any D. elongata insects are released at a site. Therefore, in a best-case scenario, it would be collected during the spring and summer of 2002 at the Canadian River site. This baseline data is a vital part of the site monitoring program and will make it possible to determine the effects of biocontrol on not just the saltcedar but on non-target vegetation and the wildlife in the area. Measurements are taken on a variety of things such as the density of saltcedar foliage and trees, variety of other plants present, what wildlife (especially birds) use the site, threatened/endangered species, and the water levels/flows in the area. The types of data gathered and how to collect the data have been decided and field-tested by the Saltcedar Biocontrol Consortium members.

Phase III. Establishment of an Initial Site for the Biocontrol Insects
The initial site chosen for establishment of biocontrol insects at the Canadian River will need to be away from any possible flooding, with a good cover of litter and a number of young trees available. Easy accessibility and security are other considerations. It is best to release the insects into a cage at first. A 12'L x 12'W x 6'H cage can usually contain several topped trees. Initially confining the insects to a cage makes it much easier to locate them and to assess how they are doing. The cage will need to be checked every week or two to be sure the insects don't run out of food. Once the insects have successfully overwintered and established, the cage can be removed.

Phase IV. Redistribution of Insects
Depending on the size of the overall area at the Canadian River and what we learn about the insects= natural dispersion rate, human intervention may not be necessary for redistribution. If redistribution is necessary, beetle populations should be allowed to build at the original release site for one or two years. Then some can be collected, either from the field or from a tent set up for breeding, and taken to other sites. The easiest life stages to move are the adults, followed by eggs; larvae are more fragile to handle.

Phase V. Continued Monitoring of Insects, Plants, and Wildlife
Monitoring will be similar to that done in Phase II, but with the addition of information about D. elongata numbers and effects. Monitoring should be continued annually for at least three years, and plans made for additional long-term checks. Some monitoring, such as insect monitoring, will be timed to occur at the same specific time or stage in the insects' life cycle each year. This monitoring may seem unnecessary, but it is actually the only way to check whether the insects are having their intended effect, or if adjustments need to be made.

Phase VI. Rehabilitation of Impacted Areas
It will take a number of years before the beetles build to populations that decrease the saltcedar cover. Mature saltcedar trees have energy reserves that allow them to resprout after a number of defoliations. Revegetation of affected areas will be important to provide wildlife habitat, avoid soil runoff into rivers, and keep other undesirable plants from becoming established.

Areas cleared of saltcedar will fall into the following categories:

Possible Advantages and Disadvantages of Using Biocontrol for Saltcedar




For extensive stands of saltcedar, such as those along the Canadian River, biocontrol is an option that should be considered. Biocontrol may not work in all areas of the Canadian River infestation, but over time could reduce the amount of saltcedar enough to improve in-stream flows. Biocontrol would provide long-term, nondisruptive, selective control of saltcedar at the site. A commitment to long-term monitoring of the insects, vegetation, and wildlife, and to any necessary revegetation of sites may be required to gain federal approval for the project. However, these requirements may not differ from those needed for approval of large-scale implementation of other control methods.


Dave Sisneros, Upper Colorado Region Saltcedar Cost Analysis/Evaluation, Applied Sciences Referral Memorandum No. 94-2-2, Feb 1994.

[Bureau of Reclamation Home Page]