Search

1 A removal experiment was carried out in a species-rich limestone grassland in Derbyshire, England. Biodegradable herbicide was used to create eight sizes of gap around in situ ramets of five co-occurring plant species: Briza media, Carex caryophyllea, Lotus corniculatus, Plantago lanceolata and Sanguisorba minor. 2 For all species, plants in larger gaps had more leaves than control plants, although there was an associated decrease in mean leaf length. This morphological response was probably due to changes in light quality. 3 Total leaf length of the target plants remained similar between treatments despite changes in leaf morphology. Flower production was also unresponsive to gap size. The biomass of the three smallest species increased with increasing gap size, largely due to higher specific shoot weight. 4 Both the importance and intensity of competition (measured, respectively, as the variation explained and slope in regressions of above-ground biomass against gap size) were greatest for the three smallest species; the biomass of both Plantago and Sanguisorba was unaffected by diffuse competition over the range of gap sizes studied. 5 Morphological plant characteristics (leaf number and mean leaf length) were more sensitive to gap size than the traits more closely related to fitness (biomass, flowering and total leaf length) and as such may be buffering changes at this higher level. Keywords: calcareous grassland, coexistence, diffuse competition, morphological plasticity, removal experiment, species-rich

1 Removal experiments are widely used to study competition in natural ecosystems but suffer from a number of methodological drawbacks. We describe two glasshouse bioassays using field soil which were designed to 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 significantly 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 infection of Holcus lanatus roots and subsequent shoot growth. In the same experiment, a nonmycorrhizal species (Cerastium fontanum) was unaffected by soil disturbance. 4 These findings suggest that the disturbance resulting from total plant removal will have adverse effects on the performance of plants remaining in the sward. Since the presence of roots seems to have little effect, above-ground clipping and leaving roots to decay in the soil is likely to be a better approach to removal experiments. Keywords: bioassay, removal experiments, soil disturbance, calcareous grassland, mycorrhizal colonization, decaying roots,