| Abstract: | Calculation of a community matrix for a lawn, based on pairwise competition experiments between lawn species, had predicted that the community would be unstable, but that it was close to the stability/instability boundary. To test for stability in the actual community, three types of 'pulse' perturbation were applied: shading for six weeks, mechanical perturbation (removal of vegetation) and a herbicide specific to grasses. Following shade and mechanical perturbation, the abundances of the species recovered towards that of the undisturbed community, although full recovery of the whole community was not realised. For the herbicide perturbation, recovery was still quite incomplete after 4.5 years, apparently because the compensatory growth of dicots made it difficult for the grasses to reinvade. Based on species presence/absence, the lawn fully recovered from all three perturbations. The overall pattern of recovery of the lawn was consistent with the prediction of marginal instability derived from the community matrix analysis. However, care must be taken in the interpretation of these results, as a pre-requisite for valid application of community matrix theory is that the community should be at equilibrium, but in the lawn this assumption was violated. A transitive hierarchy of competitive dominance was identified in the pairwise competition experiments; however, this hierarchy was unrelated to the rankings of the species in field abundance in the undisturbed community. The problems in applying community matrix theory to real communities are reviewed. It is concluded that whilst the theory may have biological relevance to at least simple communities, such as the lawn system used here, difficulties in its practical application, and more importantly, the restrictive assumptions on which the theory is based, will limit its relevance in most natural systems. These results call into question the generality of a large volume of theoretical studies based on these methods. |
