| The study of perturbed ecological communities provides a unique opportunity to observe the reassembly of the community and thereby study the evolution of biocomplexity. The ability of four measures of biocomplexity (quadratic entropy, taxonomic entropy, total complexity and total diversity) and their components to quantify the nature and degree of recovery in perturbed plant communities was assessed by how closely each measure reflected the expected trends in the evolution of biocomplexity. The measures were first applied to two plant communities, located in Sudbury, Canada and Harjavalta, Finland, recovering naturally along gradients of decreasing perturbation resulting from nickel-copper smelter emissions. The measures were also applied to the study of rehabilitated sites in the Canadian case to assess differences between assisted and unassisted recovery. Of the four measures of biocomplexity, taxonomic entropy was found to best reflect the expected trends and provide additional insight as to the nature of recovery along the gradients of perturbation. Quadratic entropy was best able to distinguish between subtle differences in recovery among assisted sites and unassisted sites. However, the taxonomic component of taxonomic entropy provided the simplest description of the underlying taxonomic structure at these sites. Taxonomic entropy was determined to be the most useful of the four measures as it has a clear information-theoretical meaning, the relative contributions of classical diversity and taxonomic information can be easily isolated and its computation is straightforward relative to total complexity and quadratic entropy. Additionally, it is relatively insensitive to community evenness and is thereby less sample size dependant and a more conservative measure to use for comparisons between studies and for conservation and restoration purposes. |
