It is hypothesized that mutualistic relationships can be persistent du
e to a third ''external'' species constraining the otherwise unlimited
growth of the mutualists. By consuming mycorrhizal fungi, soil fauna
may exert such a control over the development of the plant-fungus symb
iosis. In this study I investigated the influence of soil fauna on the
mutualistic association of ectomycorrhizal (EM) fungi and seedlings o
f birch and pine in transparent experimental microcosms simulating con
iferous forest soil. The soil was defaunated, reinoculated with soil m
icrobes, and seedlings of silver birch (Betula pendula) or Scots pine
(Pinus sylvestris) infected with 3-4 taxa of EM fungi were separately
planted in the microcosms. Half of the microcosms were thereafter rein
oculated with diverse soil fauna. Microcosms were incubated in a clima
te chamber with varying illumination and temperature regimes for two a
nnual cycles of the plant. After 57 wk the biomass of EM fungi was sig
nificantly reduced in the refaunated microcosms, the amount of fungi o
n birch and pine roots being 50 and 20%, respectively, of that in the
non-refaunated controls. Similarly, the total microbial biomass in the
bulk soil was significantly lower in the refaunated systems. The shoo
t production of both tree species was significantly enhanced in the pr
esence of fauna: by the end of the experiment 1.5 (birch) and 1.7 (pin
e) times more aboveground biomass was produced in the refaunated micro
cosms than in the controls. Concentrations of N and P in the foliage o
f both tree species were also higher in the refaunated microcosms. The
se data indicate that the obvious reduction in EM fungi due to soil fa
una is not harmful for the plant-fungus symbiosis provided that the co
mmunity composition of the belowground food web is complex enough to e
nsure efficient mobilization of nutrients in the mycorrhizosphere.