Benomyl (a fungicide) and phosphate (as NaH2PO4.cntdot.H2O) were applied to patches of two annual serpentine communities dominated by forbs. After one year neither Benomyl nor phosphate had affected plant productivity (except for legume productivity which was increased by phosphorus application), the abundance of any species or mycorrhizal infection. However, at both research sites Benomyl decreased phosphorus concentration in the shoots whereas added phosphorus increased its concentration. At one site, added phosphorus also increased nitrogen concentration in the shoot. In the following year, nitrogen (as NH4NO3) was added to the experimental design. In that second year, both nitrogen and phosphorus applications decreased the abundance and total above-ground dry weight per unit area of Plantago erecta (Morris), a dominant annual forb, and decreased the density of all plants taken together. Additionally. Benomyl increased total plant density. Phosphorus and nitrogen increased productivity of the whole community. The data presented herein support the hypothesis that communities occupying inherently infertile sites are not very responsive to nutrient amendment and that mycorrhizal infection may serve as a controller of community structure.