Community structure and assembly rules in a dune slack: Variance in richness, guild proportionality, biomass constancy and dominance/diversity relations

Abstract

There is still disagreement as to whether plant communities can be described as conforming to certain assembly rules. We searched for such rules in the vegetation of four dune slacks in west Wales by examining variance in richness, guild proportionality, biomass constancy and dominance/diversity relations. To test for the occurrence of niche limitation, variance in quadrat species-richness was compared to null models of species occurrence. Using an Overall-model, one site showed a deficit of variance, but this disappeared when a Patch-model analysis was used. It therefore seems likely that the effect observed was due to equality in the species pools adapted to different microenvironments. The proportional representation of functional guilds was examined, in guilds based on morphology and life history. Significant constancy of guild proportions was seen for the annual guild in the only site in which it occurred, perhaps reflecting a regeneration niche. Total biomass per quadrat was examined, and compared to a model in which the biomass of each species was determined at random from the biomasses of that species at the site. Surprisingly, the observed biomasses showed no departure from the null model. However, there was considerable discrepancy from an alternative null model in which biomasses were allocated at random irrespective of species. This would imply that competition limits species coexistence, rather than performance. Examination of the proportions of species from different potential-size classes in each quadrat gave limited confirmation. The local abundance of species (mean biomass) was significantly correlated with the frequency of the species within a site, parallelling the geographical-range/local-abundance correlation previously reported. Five dominance/diversity models were fitted to the biomass data. The best fits were given by the Geometric or the General Lognormal models, though which fitted best varied between and within sites. Between sites, there was some indication of a consistent difference in Preston's ‘gamma’. Most of the results are compatible with Gleason's Individualistic concept of the plant community, though there are several pointers that assembly rules have some impact on community structure.