ON DECAYING ROOTS, MYCORRHIZAL COLONIZATION AND THE DESIGN OF REMOVALEXPERIMENTS

1 Removal experiments are widely used to study competition in natural ecosystems but suffer from a number of methodological drawbacks. We de scribe two glasshouse bioassays using field soil which were designed t o test for effects of removal experiment methodology on plant growth. 2 Soil was sieved to remove plant roots which were then added back to some samples. The presence of decaying roots in the soil did not signi ficantly alter the performance of Holcus lanatus plants over the time scale of the experiment (three months). 3 Over the same period of time , however, sieving the soil significantly reduced the mycorrhizal infe ction of Holcus lanatus roots and subsequent shoot growth. In the same experiment, a nonmycorrhizal species (Cerastium fontanum) was unaffec ted by soil disturbance. 4 These findings suggest that the disturbance resulting from total plant removal will have adverse effects on the p erformance of plants remaining in the sward. Since the presence of roo ts seems to have little effect, above-ground clipping and leaving root s to decay in the soil is likely to be a better approach to removal ex periments.