A simple model is proposed which describes the relation between the extent
of mycorrhizal colonization of a plant and the 'benefit' (positive or negat
ive) derived by that plant. 'Benefit' is defined as the percent change in a
plant performance parameter of a mycorrhizal individual relative to the me
an of a number of mycorrhizal plants, grown in identical conditions. The mo
del predicts a general curvilinear relation between colonization density an
d benefit, where benefit is maximized at some value of colonization. It is
based on the fact that the relation of plant P uptake to mycorrhizal coloni
zation is often non-linear. Four examples of empirical data which provide a
good fit to the model, with third-order polynomial regression are given. I
t is suggested that if the curvilinear relation is of general occurrence th
en it can provide an explanation for many of the apparently anomalous resul
ts seen in the mycorrhizal literature.