The incidence of take-all of wheat, caused by Gaeumannomyces graminis var. tritici (Ggt), and eyespot, caused by Tapesia yallundae, was examined in a long-term rotation-tillage experiment at Wagga Wagga, N.S.W. Take-all occurred in years of higher August-October rainfall from 1979 to 1984. In years with take-all, soil water in the upper 20 cm was estimated to be above permanent wilting point for the growing season. Eyespot was associated with above-average rainfall during winter and spring and was more prevalent where residues of wheat or grasses were retained. After the severe drought of 1982, take-all developed to high levels in 1983 in wheat that followed wheat, lupins or pasture when stubble was retained, but was reduced in 1984 after lupins. Take-all was reduced in the lupin-wheat rotations by removing stubble through burning or by early incorporation of stubble. Take-all incidence was less in wheat that followed grazed pasture than after mown pasture. Where stubble was retained, Ggt survived on stubble from wheat in 1981, through the drought of 1982, to infect wheat in 1983, but inoculum did not survive on stubble through the wet season of 1983 to infect wheat in 1984. Regression analysis indicated that take-all was negatively correlated with yield but eyespot was not. Take-all reduced yield by reducing kernel mass in 1 year and by reducing kernels per cm2 in 2 other years. Soil water conditions that were associated with take-all development from 1979 to 1984 occurred in 50% of years from 1960 to 1989.
The Importance of Long-term Observations for Understanding Dryland Soil-Water-Vegetation-Climate Dynamics
