Comparison of natural and artificial epidemics of take-all in sequences ofwinter wheat crops

Winter wheat was grown for six successive years (Expt 1) and for three succ essive years (Expt 2) in field experiments on different soil types. Artific ial inoculum of the take-all fungus (Gaeumannomyces graminis var. tritici c ultured on autoclaved oat grains) was incorporated in the soil of some of t he plots just before, or at, sowing of the first winter wheat crop. Expt I tested the incorporation of similar amounts of inoculum (212 kg ha(-1)) at different depths. Expt 2 tested different amounts of inoculum at the same, shallow depth. Early sowing (September), late sowing (October) and spring i noculation were additional treatments, applied to the first crop only, in E xpt 2. Seasonal factors apart, the disease outcome in the first year after inocula tion depended an amounts and placement of applied inoculum, as well as date of sowing. Deeper inoculum resulted in less disease (Expt I). Severe take- all was produced in Expt 2 by incorporating inoculum shallowly in sufficien t quantities (400 kg ha(-1) or more). Less inoculum (200 kg ha(-1)) generat ed less disease, especially in earlier-sown plots. Differences in disease a mongst inoculum treatments were greatest in the first year and diminished s ubsequently, particularly where sowing had been early in the first year. In Expt I, where first crops exposed to artificial inoculum developed moder ate-to-severe disease, disease in subsequent second and/or third crops was less. In the fourth crop a second peak of disease occurred, coinciding with a first peak in sequences without added inoculum. Take-all decline (TAD) a ppeared to be expressed in all sequences thereafter. In Expt 2 in sequences without added inoculum, TAD occurred after a peak of disease in the second crop. Where 400 kg ha(-1) or more of inoculum were added, disease was seve re in the first year and decreased progressively in successive years. Disea se was less patchy in plots that received artificial inoculum. However, it remains uncertain that severe disease caused by artificial inoculation achi eved an early onset of true TAD. The infectivity of the top 12 cm of soil in the first 3 yr of Expt 1, deter mined by bioassay, depended on the depth of added inoculum and amount of di sease in subsequent crops. However, at the time of the naturally occurring peak of disease severity (in either inoculated or non-inoculated plots) it did not predict either disease or TAD. Differences and similarities amongst epidemics developing naturally and tho se developing from different amounts and placement of applied inoculum have been revealed. The epidemiological implications of adding inoculum and the potential value of artificially-created epidemics of take-all in field tri als are discussed.