Soil factors affecting the yield of winter wheat: analysis of results from I.C.I. Surveys 1979-80

The soils of fields where winter wheat yields were measured for the I.C.I. Ten Tonne Club Survey in 1979 and 1980 were identified according to the soil series recognized by the Soil Survey of England and Wales. The yield distributions were almost normal, with means 1·49 and 1·60 t/ha greater than the national average. Soil series accounted for 18–19% of the yield variance in both years, which was much more than any other single factor measured. If introduced in best order, soil series, crop variety, previous crop and sowing date accounted for 31% of the variance in both years. However, classes based on soil particle size and drainage, as derived from series identifications, accounted for very little of the variance. Only thick brickearth (loess) soils gave greater mean yields in both years than the overall means of all sites where series were identified. The same soil types provided 35% of the sites where > 10 t/ha was achieved in either year, though a wide range of soil types gave such large yields. Yields were greater in 1980 than in 1979 in almost all parts of the country. Although rainfall was only slightly less in the spring and early summer of 1980 than in the same period in 1979, many parts of the country suffered large potential soil moisture deficits in 1980, but these decreased yields slightly on a few series only. Factors other than rainfall used to calculate soil moisture deficits (radiation or wind) probably affected yields much more than rainfall itself.