EFFECT OF VESICULAR‐ARBUSCULAR MYCORRHIZAL FUNGI ON THE RELATIONS OF PLANT GROWTH, INTERNAL PHOSPHORUS CONCENTRATION AND SOIL PHOSPHATE ANALYSES

Summary The effect of infection by vesicular‐arbuscular mycorrhiza on the predictive value of soil analysis for available P was tested in a glasshouse experiment. Leeks ( Allium porrum ) were grown on ten Rothamsted soils with a wide range of initial NaHCO 3 ‐soluble P, each of which also received 5 levels of added P. Soils were partially sterilized with 1 Mrad of γ‐radiation, or left untreated; plants on sterilized soil were infected with yellow‐vacuolate endophyte ( Glomus mosseae ) or left non‐mycorrhizal. The NaHCO 3 ‐soluble P in the soils was measured 5 days after phosphate addition. Yields from all P levels on all soils were plotted against soil NaHCO 3 ‐soluble P content and separate smooth response curves were obtained for non‐infected and artificially‐infected plants. Infection increased yield only on soils with less than 100 mg P kg −1 . In contrast, naturally‐infected plants gave no clearly defined response curve to P, and on three soils the yields were consistently low. These soils gave very low natural infections and had low spore densities probably due to the cropping history of the sites. Artificially‐infected plants had much higher shoot P concentrations than did uninfected plants of similar dry weight, over a large range of soil phosphate levels; this effect was also noted with the endophyte G. fasciculatus . Naturally‐infected plants showed a poor relationship between shoot P concentration and yield.