Predicting community persistence based on different methods of species ranking.

Different species in a community can be ranked according to the strength of their effect on the dynamics of the entire community. Despite a considerable research effort on community structure and the “keystone” species concept, there are still unresolved issues in this area. This work addresses the suitability of different methods of characterising community members, either by relative abundance, or different measures of competitive ability. Multi-species competitive communities were simulated to allow comparison of methods that rank community members by either their abundance or competitive ability. Results indicate that characterising species according to their relative density, rather than competitive ability, provides a more refined measure of the importance of each community member. The analysis is extended by assessing which species are most likely to be involved with cascading extinction events, showing that the species with the lowest abundance is most likely to face extinction. Here it is also shown that removal of the most abundant species, i.e. that with the greatest relative density within the community, leads to the highest probability of community collapse, with further loss of community members through cascading extinctions. The asymmetry and non-linearity of between species interactions mean that species’ ranking can change unexpectedly following species loss. These results can lead to practical recommendations in management decisions for conservation and other species managementproblems, e.g. harvesting in marine food-webs, where it may be difficult to assess relative competitive ability between species in the ecosyste