Ecological Agriculture Projects Logo

EAP Publications | Virtual Library | Magazine Rack | Search | What's new

Join the Ecological  Solutions Roundtable



Biological Control of Weeds

by Professor Alan Watson Department of Plant Science


Biological control of weeds is the deliberate use of natural enemies to reduce the density of a purticular weed to a tolerable level The objective of biological weed control is not eradication but simply the reduction of the weed population to an economically low level In fact for biological control to be continuously successful, small numbers of the weed host must always be present to assured the survival of the natural enemy.|

The two most frequently cited examples of successful biological weed control are the destruction of the prickly pear cacti (Opuntia; spp.) in Australia by an imported moth [Cactoblastis cactorum) and the control of St. Johnswort Hypericum perforatum) millepertius perforé) on rangeland in California, Oregon, Washington, and British Columbia. These examples demonstrate that biological control can provide a permanent solution to serious weed problems. The first insects were introduced 50 years ago and the Australian prickly pear is still under control. St. Johnswort is presently controlled at two per cent of its former density in British Columbia.

Weeds are plants growing where they are not wanted" and most of our serious weeds have been introduced from other parts of the world. Of the 107 noxious Canadian weeds, 78 have been introduced from Europe or Asia. One of the reasons why these plants are so noxious in their new habitat is that their natural enemies are often absent. Biological control has most frequently been applied against these alien weeds and attempts are made to restore the natural control of these weed pests by introducing one or more host-specific, damaging natural enemies from the native region of the weed.

Methods Used in Biocontrol of Weeds

The classical approach to biological control of weeds involves the introduction of host-specific natural enemies of allien weeds. Recently the approaches utilized in biocontrol programs have been expanded to include two other methods:

1. augmentation of natural enemy populations; 2. application of "biological herbicides".

Augmentation includes the periodic release and/or distribution of natural enemies. Work is presently being conducted in the United States to determine the effect of mass rearing of a native moth Bactra verutana) in the laboratory and releasing it in fields of yellow nutsedge Cyperus esculentus) (souchet comestible). The application of spore suspensions of plant pathogens as 'biological herbicides' is discussed in more detail later.

Biological Control Agents

Insects have been most frequently used as biological control agents of weeds and this will likely continue. The reasons are that there have been major successes using phytophagous insects and almost all of the scientists working in biocontrol of weeds are entomologists. However, recent research has demonstrated the potential of other organisms, including plant pathogens, nematodes, and fish.

Procedures in Classical Biological Control

The first step in a biological weed control program is to determine the suitability of the weed for this approach. Not all weeds are suitable and those with the following characteristics are generally least suited for biological control:

1. Weed species which are valued in other situations are not good candidates for this approach. for example, blue weed Echlum Vulgate) (vipérine) is a serious pasture weed, but it is also a desirable honey plant. As the biocontrol agent cannot be limited in area, like chemical or mechanical treatments, 'weeds' that are of values in other situations are not suitable targets for biocontrol.

2. Weeds that are closely related to economic crops are not good candidates for this method. The closer the relationship the less possibility there is that a biotic agent could distinguish between the weed and the crop. For example, no insects have been found that attack wild oats Avena fatua) (folle avoine) and do not attack cultivated oats and our other cereals.

3. Native weed species are not generally amenable to this approach. However, the native range of an introduced allien weed provides a source from which a parasite can be introduced.

4. Weeds of cropland under intensive cultivation are generally not suited to this approach. Since the biological control agent is specific to only one weed species, little would be gained if one weed, such as lambsquarters, was controlled in the corn crop as numerous other weed species would soon occupy the available space. However, biocontrol is particularity suited to rangeland situations where a single, dominate weed species is troublesome.

5. Minor weed problems are not generally suited to this approach. The target weed should infest large areas. Since there are only four scientists in Canada working more or less full time in biological weed control, minor weeds usually cannot be considered.

6. If eradication of the weed is desired (e.g., poisonous weeds), the method is generally not applicable.

Therefore, in classical biocontrol of weeds, the ideal target weed is an aggressive, introduced weed which infests large areas of marginal land such as rangeland, pastures, and waste areas.

The second step in the program is to conduct a survey overseas to determine if there are any parasites available for introduction against the weed. A survey is also conducted in Canada to ensure that a prospective biotic agent is not already present in Canada and to determine what parts of the target weed are not attacked in Canadian populations. For example Canada thistle Cirsium arvense; (chardon des champs) is attacked by a number of seed feeders in Canada and, therefore, it would be of little value to introduce other organisms that attack the seed heads. Surveys are being conducted in Europe on the following weeds:


English name French name Scientific name
Bindweed Liseron Convolvvulus spp.
Absinthe Armoire Artemisia absinthium
Narrow leaved hawk's beard Crépis des toits Crepis tectorum
White cockle Lychnide blanche Lynchis alba
Bladder campion Silene enflée Silene cucubalus
European buckthorn Nerpun commun Rhamnus cathartica

The third step is to determine the potential effectiveness of the parasite in Canada In an attempt to eliminate ineffective agents before importation and screening tests in Canada. Insects that have been collected in certain parts of Europe may not survive under Canadian conditions.

The next and perhaps most important phase of the program is to determine the safety of the selected parasite for release in Canada. The introduced agent must not attack or damage any desirable plants in Canada. Very extensive tests are conducted in the quarantine facilities at the Agriculture

Canada Research Station at Regina, Saskatchewan. These studies involve thorough investigations of the agent's biology and its host range to demonstrate that the introduced agent is host-specific and will not become a pest of an economic crop. Some insects that are presently being screened at the Regina laboratory include:

Liothrips spp.--a sucking insect which attacks ragweed (Ambrosia artemisiifolia) (petite herbe à poux). Tephritis dilacerata--a seed-head fly attacking perennial sowthistle (Sonchus arvensis) (laiteron des champs). Argyroploce striana--a root moth attacking dandelion (Taraxacum officinale) (pissenlit).

If it is found during these tests that the candidate biological control agent is not host-specific, the organism is destroyed and the project is terminated.

After the screening tests are complete, a report is prepared for joint approval by the Canadian and U.S. governments. A report on the nematode, Paranguina picridis, which forms galls on Russian knapweed [Acroptilon (Centaurea) repens] (centaurée de Russie), was recently submitted and permission for limited field trials has been granted.

Once the agent has been determined safe and permission to release it has been given, the fifth step in the program is to establish the biocontrol agent in Canada on infestations of the target weed. This part of the program requires the co-operation of farmers. Infestations of the target weed need to be located and the release site must be maintained for some time with minimal disturbance. This stage is the most critical in the program as many agents have failed to become established because they were not adequately cared for during the initial phases of establishment in their new environment. Attempts are being made to establish insect agents on the following weeds in Canada :


Nodding thistle (chardon penché) Carduus nutans Rhinocyllus conicus (seed-head beetle)

Ceutorhynchidis horridis (root weevil)

Canada thistle (chardon des champs)(Cirsium arvense) Urophora cardui (stem gall fly)
Bull thistle (chardon vulgaire) Urophora stylata (seed-head fly)
Russian thistle (soude roulante) (Salsola pestifer) Coleophora parthenica (stem mining moth)
Diffuse and Spotted knapweed(centaurée diffuse et centaurée maculée) (Centaurea diffusa, C.Maculosa) Urophora affinis and U.quadrifasciata(seed-head flies)

Once the insect agent is established, the step is to determine if the agent is increasing and its effect on the weed. The beetle on nodding thistle and the seed-head flies on the knapweeds are well established and are rapidly increasing.

Two major projects can be considered complete in Canada. As mentioned earlier the population of St. Johnswort in British Columbia has been reduced to two per cent of its former density by two leaf-feeding Chrysolina beetles. Tansy ragwort Senecio jacobaea) (séneçon jacobée) has been successfully controlled in Nova Scotia by the defoliating cinnabar moth (Tyria jacobaeae).

From the above examples it can be seen that biological weed control programs have been primarily concerned with introduced insects as biotic agents and the approach has been directed against aggressive, alien weeds in rangeland situations. However, recent research results with other types of biocontrol agents, particularly plant pathogens, has extended the application of biological control to aquatic weeds and those of cultivated land.

Plant Pathogens As Biocontrol Agents of Weeds

Plant pathogens offer two advantages over insects as biocontrol agents of weeds: 1) they are often more host specific, and 2)they can be applied with conventional spray equipment at a time when the weed is at its most susceptible stage.

A. Conventional approach

As discussed above, biological control of weeds has conventionally been applied against alien weeds by introducing one or more of their natural enemies. Control of a weed with a plant pathogen used in this manner was achieved against skeleton weed (Chondrilla juncea), the major weed of wheat in Australia. A host-specific rust, Puccinia chondrillina, was collected in Italy and introduced into Australia in 1971. The rust spread very rapidly and at present controls the weed over most of its range. Control has been so spectacular that herbicides are no longer being used to control this weed in Australia.

Numerous introduced weeds in Canada are known to be attacked by plant pathogens in their native ranges. However, research in this area has not yet been established in Canada.

B. Biological Herbicides

A new approach to biocontrol of weeds has recently been pioneered in Arkansas with the application of a spore suspension of an endemic fungus. The fungus controlled 99 per cent of a serious weed (Northern jointvetch) in rice, without damage to the rice crop. The disease was already present in the fields but did not damage its host because it normally attacked too late in the season. However, when the fungus was applied early (in the seedling stage), it destroyed the weed. The fungus is now being tested on a field scale and there appears to be no biological or technological limitations that would prevent commercialization of the fungus as a' mycoherbicide'.

Possible targets for this approach in Canada are numerous as practically all Canadian weeds have plant pathogens recorded on them. for example, wild oats (Avena fatua) (folle avoine) and other weeds of our cultivated land can be controlled to some extent by cultural and mechanical means. However, they still remain as serious contaminants of our crops. Wild oats is attacked by numerous plant pathogens which normally attack late in the season and do not appreciably damage the weed. If host-specific pathogens of wild oats were applied early in the season, some degree of control could possibly be achieved.

An attractive feature of this method is that it does not involve the introduction of any new organism into our environment but merely the use of host-specific pathogens at a time when they are most effective. This method may also prove to be valuable for integrated weed control systems. For example, most weeds are controlled in corn with standard herbicide application, but a few weeds such as quackgrass (Agropyron repens) (chiendent) and yellow nutsedge (Cyperus esculentus (souchet comestible) are more difficult to control. The application of host-specific pathogens could be integrated into the present weed control system to control these types of weeds.

Aquatic Weed Control

Biocontrol of water weeds is being investigated in Florida and other parts of the U.S. The possibilities of using insects, plant pathogens, and a fish, the white amur (Ctenopharynogondon idella) are being studied. Water weeds are also a serious problem in many areas of Canada and the developments in biocontrol are being carefully observed.

The disadvantages of using herbicides in water systems are numerous and alternative methods should receive high priority. There is considerable controversy concerning the use of the fish for biological control. Research is continuing and further studies should demonstrate that the advantages will by far outweigh any possible disadvantages.

Biological Weed Control Program at Macdonald College

The research in biological weed control at Macdonald College is composed of two segments: 1. The release of screened biotic agents on infestations of weeds in Quebec; 2. Investigation of the potential of native natural enemies of certain weeds in Quebec.

A. Release of Biotic Agents

Insects that have been screened and are available for release are shipped to Macdonald College from the Regina laboratory. These insects are then taken to desirable field release sites and liberated on the target weed. This program has involved the co-operation of numerous Quebec farmers and its continued success depends on their support. Last year, insects were released on the following weeds in Quebec


Canada thistle (chardon des champs)(Cirsium arvense) Urophora cardui (stem gall fly)
Bull thistle (chardon vulgaire)(Cirsium vulgare) Urophora stylata (seed -head fly)
Leafy spurge (euphorbe ésule)(Euphorbia esula) Hyles euphorbiae (defoliating moth)

Additional releases of these insects will be made this year and releases of Chrysolina beetles on St. Johnswort, Tyria.Jacobeae (cinnabar moth] on tansy ragwort and Uronhora spp on spotted knapweed infestations in Quebec are planned for this year.

Urophora cardui has previously been released at St. Hyacinthe and Rhinocyllus conicus has previously been established on nodding thistle in the Lac St. Jean region.

B. Native Natural Enemies

This program involves surveying of weed populations, collection of natural enemies (plant pathogens and insects) and determination of the potential of selected organisms as biocontrol agents. At present, two post-graduate students, Jean-Guy Champagne and Harry Hartmann, are investigating the host-parasite relationships of Canada thistle and its rust pathogen and ragweed and its natural enemies. In addition endemic plant pathogens of other weeds, such as quackgrass and yellow nutsedge, are also being collected and studied.


Biological weed control has recently received renewed interest because it is an environmentally compatible method of weed control without residue and pollution problems. However, biological weed control has its weaknesses and is not suitable for all weed problems. Indeed, only a few of our many noxious weeds have been investigated for biocontrol and so far only two are sufficiently advanced to be considered complete successes. Undoubtedly, with continued interest and research support, the general assumption that biological weed control is used only as a last resort where chemical and other methods have failed will be proven incorrect.

Copyright © 1977