Abstract

Take-all is a root disease of wheat caused by the fungus Gaeumannomyces graminis var. tritici (Ggt). The most common method of control, growing wheat after a break crop, is not always feasible. This study compared the use of a break crop with 5 alternative control methods in a series of field experiments in south-eastern Australia. The methods of control tested were: (1) fungicide added to fertiliser; (2) soil fumigation with methyl bromide; (3) applied chloride; (4) seed treatment with microbial antagonists; (5) a prior brassica break crop; and (6) a 12-month-long fallow. Eight experiments were conducted over 2 years but not all treatments were included in each experiment.

The most successful control methods were growing wheat after a brassica break crop or a long fallow. Both methods gave 72% yield increases over wheat growing after wheat. None of the other methods gave consistent, significant, or profitable yield increases or disease control. The mean yield increases in the year of application were 8% for the fungicide, 6% for microbial antagonists, 4% for chloride, and 7% for fumigation. The probable reason that fungicide and microbial antagonists were ineffective was that they were localised in the furrow where they were applied, whereas roots became infected in the inter-row space. Probable reasons that chloride was ineffective were that the background soil chloride levels were generally above the responsive range, and that roots became infected with take-all after the chloride was leached from the topsoil. The limitation of fumigation was that it suppressed natural antagonists of the Ggt, apparently leading to reinfection at higher levels than before. There was also evidence of Ggt re-infection in the second year after break crops, leading to an apparent ‘boomerang’ effect.

Take-all inocula at the sites were measured in pre-sowing soil bioassays, whereas disease incidence was determined in seedlings and as ‘whiteheads’ as crops approached maturity. The only consistent pattern among the measurements was low disease incidence after break crops and the long fallow. Otherwise, there were low correlations between the 3 sets of measurements, suggesting that environmental changes after the soil bioassay and seedling assessment played critical roles in the progress of the disease.