Department of Environmental Biology, University of Guelph, Guelph, Ontario Nl G 2Wl, Canada

Background and objectives
Increasing public concern over the reliant use of chemical pesticides to control weeds such as dandelion, Taraxacurn officinale Weber, is one issue that has stimulated the development of biological control agents [1]. A collaborative university-industry research project was established in Canada to develop a biological control agent for dandelion in turfgrass. As part of the formulation work for this project, sublethal herbicide treatments were combined with pathogens to improve weed control and decrease costs. The objectives of this study were to: (i) evaluate fungitoxicity of three herbicides on three dandelion pathogens; and (ii) evaluate sequential treatments of sublethal herbicide rates and pathogens for injury to dandelion in growthroom and field conditions.

Materials and methods
To determine fungitoxicity of the herbicides on the pathogens, radial growth of colonies grown on herbicide-amended media was measured. The pathogen treatments included isolates G5/2 (Phoma herbarum), MAC2 (Curvularia inaequalis), and AC9530 (Plectosphaerella cucumerina). Glyphosate (0.0 to 3000.0 p.p.m.), 2,4-D (0.0 to 2800.0 p.p.m.), and diquat (0.0 to 890.0 p.p.m.) were individually diluted in autoclaved potato dextrose agar and poured into petri dishes. After 24 h, actively-growing plugs of each fungus were inoculated centrally on the media. Radial growth was measured at 6 days after inoculation (DAI) for G5/2, 8 DAI for MAC2, and 10 DAI for AC9530. For growth-room experiments, suspensions of selected fungi were applied to dandelions 30 min after treatment with diquat, and 4, 8, or 16 days after treatment (DAT) with glyphosate or 2,4-D. Up to five rates of each herbicide were selected for individual and combination treatments. Fungal suspensions were formulated as 20% (v/v) blended mycelia in potato dextrose broth (PDB). Controls were sprayed with water and PDB. After inoculation, dandelions were placed in a mist chamber for 15 h. Dandelion injury was visually assessed at 7, 14 and 21 DAI using a biomass reduction scale with rating levels from 0 to 10 (0=0 to 10% biomass reduction compared to the control, 1=11-20%, ..., 9=91-99%, 10=100%). Roots and shoots were harvested, dried, and weighed at 21 DAI. Field experiments were established in 1996 and 1997 in Guelph, Ontario, Canada. Dandelions were grown in the greenhouse for 4 weeks, transplanted to the field, and acclimatized for 2 weeks before treatment. Plots consisted of 10 plants in a 1 m2 turfgrass sward consisting primarily of Kentucky Bluegrass (Poa pratensis). Treatments were arranged in a split-plot design. Plants were assessed at 7, 14, 28, and 42 DAI. Shoots were harvested, dried and weighed at 42 DAI.

Results and conclusions
Isolate AC9530 showed little response to either diquat or 2,4-D concentrations, however, at the highest concentration of glyphosate radial growth was reduced to half that of the control treatment. Radial growth of G5/2 and MAC2 decreased as herbicide concentrations increased. The results suggest that certain herbicide concentrations may prove useful for tank-mixes with pathogens. Sequential applications of herbicide and pathogen identified synergism in some combination treatments in growth-room studies. For example, sequential treatment with diquat (1.6-3 kg ae/ha) and AC9530 provided significantly (P<0.05) better control of dandelion than either treatment alone or the sum of the individual treatments. In 1996 field studies, environmental conditions were not conducive for disease, and no disease occurred in plots treated only with pathogens. In 1997 field studies, disease was more severe, resulting in low ratings and significant (P<0.05) interactions between pathogen and herbicide treatments. Sublethal herbicide rates combined with fungal pathogens may create a synergistic interaction, potentially increasing weed control and reducing costs sufficiently to permit commercial production. Continued development of biological agents supplemented with low-rate herbicide treatments may reduce concerns of highvolume chemical pesticide use while providing effective weed control products.

1. Riddle GE, Burpee LL, and Boland GJ, 1991. Weed Science, 39,109-118.