Parsimony analysis and ecological communities, phytosociology, nested subsets analysis, forest fragmentation: a reply...

We suggested using parsimony analysis to study community evolution in terms of species composition and to apply these results in the context of forest fragmentation as a replacement for the so-called "nested subsets analysis" or other phenetic synecological or phytosociological methods (Pellens et al., 2005). Giannini and Keller (2007) took issue with this new application on the basis of three misunderstandings. We re-emphasize that communities themselves are analyzed, not landscape parts such as forest fragments. Therefore, it must be clear that communities are analogous to taxa and landscape parts such as fragments are analogous to distributions of taxa. Community evolution is the change in community composition by immigration, emigration and local extinction. Thus, gains and losses of species should not be confused with horizontal transfer. Parsimony analysis does not necessarily group communities based on shared absences of rare species. Rare species are not necessarily absent in the same communities and these absences are not necessarily inferred to be synapomorphies after rooting. This is the main advance expected when cladistics is used instead of the previously cited phenetic methods working with overall similarity.