The susceptibility of roots to infection by an arbuscular mycorrhizal fungus in relation to age and phosphorus supply

An apparatus in which plant roots may be challenged uniformly with inoculum of arbuscular mycorrhizal fungi is described. Seedlings of leek (Allium porrum L.) or clover (Trifolium repens L.) were first grown non-symbiotically in the apparatus for 21 d at three rates of phosphorus (P) addition to soil (1 50 (PI), 450 (P3) and 750 (P5) mg P kg-1 soil). The positions of individual root tips were recorded, and the root systems then challenged with inoculum of Glomus mosseae (Nicol & Gerd.) Gerdemann & Trappe. Roots were excised 14 d later, and the probability of occurrence of internal infection in successive 3 mm (clover) or 5 mm (leek) sections of root was estimated in first-order laterals (clover) or main axes (leek) from the proportion of sections at each location of replicate roots that bore internal fungal structures. Only in the region of a root proximal to the position of the root tips at inoculation could data be used to investigate change of probability of infection with cell age. Here, there were sharp declines in probability of infection with proximal distance, in both hosts and in all P treatments. The decline of probability was greater in clover: when expressed in terms of cell age at the time of challenge, there was no infection at Pl in cells > 10 d old in leek and none in cells > 7 d old in clover. Models of the form log(e) [p(i)/(1-p(i))] = alpha + beta x distance, where p(i) is the estimated probability of infection and alpha and beta are constants, were fitted to these data. The odds on infection are [p(i)/(1-p(i))]. For leek, 8 was unaltered by P addition (P3 and P5 curves were parallel to P1) but from alpha it could be calculated that on average the odds on successful infection at any particular distance were reduced by 3 7 % and 70 % by P3 and P5 rates of P addition respectively. In clover the curves for the three P treatments were not parallel. Addition of P appeared to reduce the odds on infection of clover much more than those of leek. We conclude that the simplest explanation for the patterns of infection in leek is that P addition increased the time taken for soil inoculum of G. mosseae to infect roots: the mechanism in clover might be more complex.