Methods:

The first objective will address the development of bioherbicide agents and a bioherbicide mixture to control several highly important weeds including nutsedges, pigweeds, grasses, purslanes, spurges, kudzu, and others. The biocontrol agents proposed here have all been previously shown to be effective in weed control and are critically in need of further research and development to realize their registration and utilization.

Dactylaria higginsii, recently isolated from purple nutsedge in Florida, has been shown to have good potential as a bioherbicide for this weed. When applied as a postemergent foliar spray, it causes high levels of shoot mortality in a number of nutsedge species. (Kadir and Charudattan, 2000). The host range of the fungus is restricted to Cyperus spp. and Kyllinga brevifolia (=Cyperus brevifolius), and it has been shown to be capable of suppressing purple nutsedge competition with tomato and pepper (Kadir and Charudattan, 2000; Kadir et al., 1999, 2000a, b). The bioherbicidal use of this fungus has been patented (Kadir and Charudattan, 1997), and industrial collaboration has been set up by the University of Florida to develop and register this bioherbicide.

Roles and responsibilities: the following stations will participate in this objective: USDA-ARS-SWSRU, MS; Sylvan, PA; FL SAES; IN SAES; CA SAES, Riverside; NC SAES; MT SAES; USDA-ARS-USHRL, FL; and PR SAES. Individuals from these stations (See Appendix-Projected Participation) will follow the common protocols described below.

Experimental and commercial prototype inoculum preparations of D. higginsii will be provided to cooperators. The fungus will be field-tested in different crops chosen by the cooperators according to regional needs. The potential to use D. higginsii as an alternative to methyl bromide for nutsedge control will be determined in some states. The primary target will be purple nutsedge. Yellow nutsedge, green kyllinga, rice flat sedge (Cyperus iria), and globe sedge (C. globulosus) may also be targeted. Tests will be done using common experimental protocols and experiments will be replicated and repeated. Data gathered will include percent weed control, aboveground biomass, tuber number, and tuber weight. Effects of D. higginsii on the competitiveness of nutsedges and on the growth and yield of selected crops will be studied. Dactylaria higginsii will be applied as a post-emergence over-spray in one half of the treated plots. Experiments will be repeated. Data taken will include number of nutsedge plants, percent nutsedge control, tuber number, above ground biomass, tuber weight, and crop yields (where applicable).

1-B. Development of Microsphaeropsis amaranthi and Phomopsis amaranthicola as broad-spectrum bioherbicides to control pigweeds:

Two fungal pathogens, Microsphaeropsis amaranthi (=Aposphaeria amaranthi) (Mintz et al., 1992) and Phomopsis amaranthicola (Rosskopf, 1997; Rosskopf et al., 2000) have been shown to be potential bioherbicide agents. These pathogens have overlapping and complementary host ranges, each being capable of controlling some, but not all weedy species of Amaranthus. The epidemiological conditions conducive for disease development and weed control by these fungi have been determined (Mintz et al., 1992; Rosskopf, 1997; Rosskopf et al., 2000). The nontarget safety of these fungi has also been confirmed: all pigweed species screened were susceptible, whereas the crop plants tested were immune (Mintz, 1991; Rosskopf, 1997). The bioherbicidal use of P. amaranthicola has been patented (Charudattan et al., 1995,1996).

Roles and responsibilities: the following stations will participate in this objective: USDA-ARS-SWSRU, MS; FL SAES; IN SAES; NC SAES; MT SAES; USDA-ARS-USHRL, FL; and PR SAES. Individuals from these stations (See Appendix-Projected Participation) will follow the common protocols described below.

Preliminary unpublished data on the complementary interactions between M. amaranthi and P. amaranthicola indicate that the two fungi may be used in combination to obtain broad-spectrum control of several species of Amaranthus and increase the level of weed control through synergism between the pathogens. Further multistate research will be key to the next phase of technology transfer with both fungi. Accordingly, we propose to develop methods for large-scale production and bioherbicide formulations of these fungi, evaluate their efficacy in commercial fields, and establish systems to integrate the bioherbicides in crop production.

Large-scale production of these fungi will be developed in the participants' laboratories and transferred to an industrial partner. Various materials and culturing methods will be evaluated for large-scale production. Solid-state and liquid-fermentation methods will be evaluated, by varying factors such as carbon:nitrogen (C:N) ratios, types of C and N, temperature, light, aeration, and the addition of solid material such as wheat husks, corn-cob grits, and others. Suitable prototype commercial preparations of these fungi will be produced and tested in multistate field trials using different crops. Testing will be done using common experimental protocols and the studies will be replicated and repeated. Percent weed control and crop yield will be the criteria used to determine the efficacy of these bioherbicide agents.

Weedy grasses are difficult to control. Depending on the location and crops, different grass species cause problems. Chemical herbicides and mowing are commonly used in appropriate situations, but these methods are costly and may have adverse environmental impacts (e.g., pesticide runoff from lawns and golf courses). Furthermore, grass herbicides, such as atrazine, metribuzin, and simazine, are at risk of being phased out as a result of the FQPA (Anonymous, 2001). Biological control offers an alternative, and Chandramohan et al., (2000) have shown that Drechslera gigantea isolated from large crabgrass (Digitaria sanguinalis) and Exserohilum rostratum and E. longirostratum, respectively from johnsongrass (Sorghum halepense) and crowfootgrass (Eleucine indica) can be used as a cocktail to control seven or eight different weedy grasses. In a field trial conducted in Lake Alfred, FL, crowfootgrass, guineagrass (Panicum maximum), johnsongrass, large crabgrass, southern sandbur (Cenchrus echinatus), Texas panicum (Panicum texanum), and yellow foxtail (Setaria glauca) were controlled with the mixture of these fungi. In another field trial conducted in Ft. Pierce, FL, the fungal mixture was effective in controlling a natural stand of guineagrass. When applied in an invert emulsion, two applications of these pathogens gave 85% control of all seven grasses in two weeks after the second application, and the control lasted for 10 to 14 weeks without regrowth of the grasses (Chandramohan et al., 2000). The use of a mixture of two pathogens for simultaneous control of two weeds has been demonstrated previously (Boyette et al., 1979), but this system has not been commercially developed and subjected to the EPA registration process. Therefore, a multistate effort will be undertaken to test the pathogen cocktail under different climatic conditions, against different grass weeds, and in different crop and noncrop situations. The results of these studies will be used to further develop the multiple-pathogen system for registration and commercialization.

Roles and responsibilities: the following stations will participate in this objective: USDA-ARS-SWSRU, MS; Agric. & Agri-Food Canada, SK; FL SAES; IN SAES; MT SAES; and USDA-ARS-USHRL, FL. Individuals from these stations (See Appendix-Projected Participation) will follow the common protocols described below.

Under this multistate project, methods will be developed for large-scale production of inoculum of D. gigantea, E. longirostratum, and E. rostratum. Nutrient requirements (C- and N-sources, minerals, growth factors) and physical conditions of culture (temperature, pH, agitation-aeration) will be studied. Efforts will be made to replace commercial laboratory media with less expensive C- and N-sources (e.g., molasses, grain mash, fish meal, soybean meal, whey, and others). Production will be attempted in stationary liquid cultures, shake cultures and small- (500 ml) and large-scale (>25 liters) liquid-fermenter cultures.

Efficacy determinations will be made in field plots. Different grasses will be targeted for control, depending on regional needs. Sugarcane, citrus, and other tree crops, turf, home gardens, and other sites where grasses are a problem will be targeted. Common experimental protocols will be developed and used among the cooperating states. Efficacy data from the above trials will be used to further develop these fungi as a bioherbicide cocktail. An industrial partner will be identified for commercial development and registration.

Common purslane (Portulaca oleracea), horse purslane (P. portulacastrum), spotted spurge (Euphorbia maculata), prostrate spurge (E. prostrata), and kudzu (Pueraria lobata) are serious weed pests in many areas of the United States, including the Southern Region. Purslanes and spurges cause problems in vegetables such as tomato and pepper. Kudzu is a notorious weed along roadsides, natural areas, right-of-ways, etc. in the South. These weeds often form dense, complex, difficult-to-control populations that cannot be easily managed with chemical herbicides. Methyl bromide has long been used as a fumigant to control purslanes and spurges, and the loss of this broad-spectrum fumigant will cause serious problems to the growers. A plant pathogen, Myrothecium verrucaria, shows promise as a bioherbicide for controlling several of weeds including sicklepod (Senna obtusifolia) and hemp sesbania (Sesbania exaltata) in addition to those named above (Walker and Tilley, 1997; Boyette et al., 1999).

When M. verrucaria spores were applied with 0.05% surfactant (Silwet L-77), the fungus caused foliar necrosis and biomass reduction in several crop, ornamental, and weed species. Plant responses ranged from no reduction in dry weight to 100% mortality. When the fungus was applied by directed spray to mixed plantings of sicklepod and soybean seedlings, with the latter mostly protected from the spray, only sicklepod plants were killed (Walker and Tilley, 1997). This capacity of M. verrucaria to cause necrotic damage on several plant species, including many weeds, provides an opportunity to develop this fungus as a broad-spectrum bioherbicide.

Roles and responsibilities: the following stations will participate in this objective: USDA-ARS-SWSRU, MS; USDA-ARS-FDWSU, MD; USDA-Forrest Service, GA; FL SAES; IN SAES; MT SAES; USDA-ARS-EBCL, France; Forestry Canada, BC; and PR SAES. Individuals from these stations (See Appendix-Projected Participation) will follow the common protocols described below.

Myrothecium verrucaria has a number of desirable characteristics that favor its use as a bioherbicide. As reported by Yang and Jong (1995), M. verrucaria is usually a weak pathogen, but severe plant damage can result when high numbers of spores are applied with a suitable surfactant. Secondary spread of the pathogen to uninoculated plants is not a problem since this pathogen does not readily reproduce on the infected host (Walker and Tilley, 1997). Therefore, additional research is justified to assess the potential of M. verrucaria as a bioherbicide.

The aim of this sub-objective is to develop M. verrucaria as a broad-spectrum bioherbicide for species of purslane, spurges, kudzu, and other difficult-to-control weeds. Under this sub-objective, the effectiveness of M. verrucaria as a bioherbicide will be determined in greenhouse and field trials. Systems will be developed for mass production of inoculum for cooperative multistate trials that will be conducted by using common protocols. Suitable experimental designs, replications, and repetitions will be included. Different crops (purslanes, spurges, and other weeds in row crops) and weed-infested sites (kudzu) will be used for field trials.

Using optimum conditions for disease development, the effect of M. verrucaria will be determined on the basis of weed kill, weed biomass, reproduction, and survival of weed propagules. Disease/damage incidence, disease/damage severity and weed control will be rated at suitable intervals. Plants will be harvested, also after suitable periods, for biomass and productivity measurements. A rating system will be used in choosing the most promising biocontrol agents.

To confirm efficacy of M. verrucaria under field conditions, field plots will be established in different states. Results from these studies will be used to further develop and register M. verrucaria as a broad-spectrum bioherbicide. Appropriate industrial collaboration will be established for commercial development and registration of this agent.

The potential to control invasive vines, such as old-world climbing fern (Lygodium japonicum), skunkvine (Paederia foetida), and others with M. verrucaria will be determined in FL.

The types and nature of toxic metabolites produced by M. verrucaria and their safety to human and environmental health will be determined.

The bacterial pathogen Pseudomonas syringae pv. tagetis (PST) is the causal agent of a disease characterized by apical chlorosis on several members of Asteraceae. Johnson et al. (1996) have demonstrated that this bacterium can be used to control various weeds in and outside the Asteraceae family. Spray application of this bacterium, at 5 x 10⁸ cells per ml, in an aqueous buffer containing the surfactants Silwet L-77 (0.1%) or Silwet 408 (0.2%) resulted in 100% disease incidence and a higher level of disease severity on Canada thistle (Cirsium arvense) than observed under natural infections. In addition to C. arvense, the following plants were severely diseased when sprayed with the bacterial cells: Ambrosia artemisiifolia, Helianthus annuus, H. tuberosus, and Tagetes erecta. Under field conditions, high levels of plant mortality (57-100%) were seen in the case of A. artemisiifolia, C. arvense, Conyza canadensis, Lactuca serriola, and Xanthium strumarium. In addition, severe injury was seen on infected Setaria viridis and Abutilon theophrasti. Symptoms appear in some species within 3 to 4 days and populations of A. artemisiifolia, C. canadensis, L. serriola, and X. strumarium were virtually eliminated, while populations of C. arvense were significantly reduced compared to controls. Tissue formed before the bacterium was applied, such as mature leaves, was not affected. However, once apical chlorosis was induced, seed production appeared to be inhibited in the case of C. arvense (Johnson et al., 1996).

The PST strains currently in use come from Minnesota or bordering states. Strains of this bacterium from other regions have not been compared for their ability to induce disease. However, white chlorosis is a symptom commonly observed in Canada thistle and a few other native composite plants across the United States. Although it has not been definitely determined, because of the symptoms and the plant family affected, it is generally assumed that the causal agent is PST. There are several compelling reasons that the correct identification of this chlorosis-inducing organism be determined. First, demonstrating that PST is already endemic to the areas where the biological control agent is intended for release would facilitate getting permits to release the bacterium as a biological control agent. Second, current studies with PST have been limited to one or a few strains from Minnesota and bordering states, while white chlorosis in Canada thistle has been reported from as far east as Maryland. The virulence of strains from other areas should be compared to insure that the best possible candidate is used when developing the final product. Lastly, genetic differences between the current test strains and endemic strains of PST should be documented so that studies on the stability and spread of the biological control agent are not confounded by observations of the endemic strains of PST.

PST is a weak pathogen of many Asteraceae plants, such as marigold. If inadvertently applied to nontarget plants, those plants could become diseased. Therefore, host range of this bacterium may be tested in greenhouses in different states, according to local needs. Test plants will be screened at various developmental stages. Preliminary greenhouse screening of PST pathogenicity on several aster weeds of turf and nursery crops is already underway. Species tested and found to be susceptible are: T. officinale, Artemisia vulgaris (mugwort), Bellis perennis (English daisy), Chrysanthemum leucanthemum (oxeye daisy), Achillia millefolium (yarrow), Ambrosia artemisiifolia (common ragweed), A. trifida (giant ragweed), Cirsium vulgare (bull thistle), Conyza Canadensis (horseweed), Galinsoga ciliata (galinsoga), Hieracium spp. (hawkweed), Senecio vulgaris (common groundsel), Solidago spp. (goldenrod), and Youngia japonica (Asiatic hawksbeard).

Field evaluations of this bacterial pathogen in Minnesota have indicated that it can effectively control Canada thistle in soybean and corn without causing crop damage. Although Canada thistle does not exist in much of the Southern Region, several other members of the Asteraceae family that cause problems in this region. Propagules of weeds of interest will be collected locally in cooperating states. Tests will be done in crop and noncrop situations.

Roles and responsibilities: the following stations will participate in this objective: USDA-ARS-SWSRU, MS; Agric & Agri-Food Canada, SK; FL SAES; MA SAES; IN SAES; USDA-ARS-SASL, MD; NC SAES; and MT SAES. Individuals from these stations (See Appendix-Projected Participation) will follow the common protocols described below.

Under this sub-objective, it is proposed to develop and register this bacterium as a bioherbicide for several weeds in the Asteraceae and a few other families. The bacterium will be provided for multistate trials. Weeds in turf, roadside, right-of-way, and natural areas will be targeted, depending on the cooperating states. The exotic invasive weed, Asiatic hawksbeard (Youngia japonica, Asteraceae) as well as Taraxacum officinale (dandelion, Asteraceae), Gnaphalium purpureum (purple cudweed, Asteraceae), Asperugo procumbens (catchweed, Boraginaceae), Galium spp. (bedstraw, Rubiaceae), Stellaria media (common chickweed, Caryophyllaceae), Ambrosia artemisiifolia (common ragweed, Asteraceae), and Bidens spp. (beggartick, Asteraceae) will also be targeted.

The causal agent of chlorosis in Composite weeds and other native plants in the various states where this type of symptom has been observed will be isolated, identified, and tested. Sites where chlorotic Composite weed and native species are present will be recorded and plant samples collected. Using standard laboratory procedures, bacteria will be isolated from the plant material showing lesions and/or chlorosis. In some cases, such as with Canada thistle, isolation of the organism may be complicated because tagetitoxin is readily transported within plants to meristematic tissues and Canada thistle plants are often interconnected through a lateral root system. Consequently, the infection may be localized in one plant while chlorosis develops in other connected plants. In such situations, it may be necessary to isolate and observe the plants before attempting to isolate the bacterium. Isolates will be bioassayed for toxin production using sunflower seedlings and seedlings of the plant species from which the organism was isolated. Using current genomic techniques, such as amplified restriction length polymorphism (AFLP), isolates will be genetically characterized and compared with other know strains of PST. For those instances where a chlorosis-inducing organism is not isolated from the plants expressing the PST-like disease, a simple technique for determining the presence of tagetitoxin in white-chlorotic tissue will be developed to determine if PST is the likely causal agent.