Rhizosphere and root-infecting fungi and the design of ecological field experiments.

As part of a wider study into the role of soil fungi in the ecology of the winter annual grass, Vulpia ciliata ssp. ambigua (Le Gall) Stace & Auquier, we applied the fungicides benomyl and prochloraz to three natural populations of the grass growing in East anglia, United Kingdom. The rhizosphere and rootinfecting fungi associated with the three populations were analysed each month between February and May 1992 when plants set seed. There were marked differences between the fungal floras associated with each of the three populations of V. ciliata, despite the fact that associated plant species and soil nutrient status were broadly similar between sites. This was attributed to wide differences in soil pH between the three populations. Prochloraz did not affect fungal abundance, but benomyl decreased the isolation frequencies of Fusarium oxysporum from roots and the frequencies of Penicillium and Trichoderma spp. isolated from rhizosphere soil, and increased the frequency of isolation of Mucor hiemalis from the rhizosphere of V. ciliata. There were also significant increases in the isolation frequencies of F. oxysporum from roots and M. hiemalis, Trichoderma spp. and Phoma fimeti from the rhizosphere of V. ciliata as plants matured. The significance of these results for the design of ecological field experiments are discussed in light of a previous study which has shown that asymptomatic root-infecting fungi can affect plant fecundity and hence abundance in natural populations of V. ciliata. We propose that differences in microbial communities between sites, controlled in part by soil chemistry, are a major factor determining plant performance under field conditions.