Take-all, Gaeumannomyces graminis var. tritici, and yield of wheat grown after ley and arable rotations in relation to the occurrence of Phialophora radicicola var. graminicola

Abstract

SUMMARYThe Rothamsted ley–arable experiments were on two fields with similar soils but with contrasting previous cropping: old grass on Highfield, old arable on Fosters field. Damage by take-all (Qaeumannomyces graminis var. tritici) occurred sooner in successive wheat crops grown after a lucerne ley and arable sequence (LU) than after a grass-clover ley and arable sequence (LC). On Highfield the difference was consistent and large, it occurred as soon as a second wheat crop was grown and resulted in wheat yielding 1 t/ha less after the LU than after the LC sequence. This difference did not persist in the next wheat crop where take-all was prevalent after both sequences. On Fosters field take-all developed more slowly and differences between sequences were mostly smaller.Wheat seedling bio-assay of soil cores from the LU and LC sequences showed that little take-all fungus persisted through the leys and that soils were much infested after a first wheat crop in the LU sequence on Highfield, but not in the LC sequence on Highfield or in either sequence on Fosters field. Microscopic examination of roots from assay seedlings and from field plants showed that Phialophora radicicola var. graminicola (PRG) was most common in soils where take-all developed slowly, but our results did not show if this was a causal relationship. The occurrence of much PRG in the LU sequence on Fosters conflicts with previous reports which associate large populations of this fungus only with grassland soils.