Effect of certain herbicide treatments on pasture composition and inoculum of the take-all fungus

The improvement of pastures by the use of a range of herbicides to eliminate grasses, and their effect on populations of the take-all fungus (Gaeumannomyces graminis vartritici=Ggt) were studied in the field (at Esperance Downs, on the south-coast of Western Australia) from 1982 to 1985. Field trials were conducted to evaluate three herbicide treatments (2,4-D amine+propyzamide; 2,4-D amine+paraquat; paraquat/ diquat) and an unsprayed control. A pot trial involving these treatments with two levels of nitrogen was undertaken to confirm treatment effects observed in the field trial. All herbicide treatments resulted in reduced grass composition of pastures, in both the year of spraying and in the second year of pasture, but reduced dry matter production in the year of spraying. In the year of spraying, however, inoculum ofGgt was reduced (P<0.1) only following the 2,4-D amine+propyzamide treatment and was greater (P<0.1) after 2,4-D amine+paraquat treatment than the unsprayed treatment. Despite reduced grass levels in the herbicide-treated plots in the second year of pasture,Ggt inoculum did not differ between treatments, nor did it after a wheat crop which followed a second year pasture. There was high correlation (P<0.001) between disease levels and dry weights of grasses in the pot trial. There was significantly less (P<0.001) grass in pots treated with herbicides compared to the unsprayed control but no difference (P>0.05) was evident between treatments. Inoculum levels were lower (P<0.05) in the treated pots than the unsprayed control with no evidence of differences among treatments (P>0.05). Nitrogen level had no effect on disease (P>0.05). All herbicide treatments tested reduced grass level and total dry matter, both in the field and in pots. Whereas in the pot trial reduced grass levels resulted in reducedGgt inoculum, in the field such a reduction occurred only with the 2,4-D amine+propyzamide treatment and only in the year of spraying. Herbicide treatments had no effect onGgt inoculum in second year of pasture or crop. Unknown soil and environmental factors in the field precluded a simple relationship between grass level in pasture and subsequent level ofGgt inoculum, and where such a relationship did occur (2,4-D amine+propyzamide treatment) it appeared to be shortlived.