Quantitative aspects of community structure examined in a semi‐arid grassland

1 Community structure was examined by analysing the spatial distribution of species' biomass–abundance, as well as species' occurrences, in two semi‐arid grassland sites in New Zealand. This is the first attempt to search for a range of assembly rules (variance in species occurrence, guild proportionality and intrinsic guilds) using quantitative data. 2 One site, with a greater proportion of native species, had a larger species pool, greater species richness and greater biomass. At this site, variance in richness was generally greater than expected at random, but variance in biomass tended to be less than expected at random, probably indicating competition. Microsites which had lower biomass and fewer species tended to be dominated by exotic species; it is suggested that these are more recently disturbed microsites. 3 At a second site, dominated by exotic species, there was no significant evidence for community structure, i.e. no non‐random patterns in richness or biomass, i.e. no assembly rules. 4 Guild proportionality was not seen in a priori guilds: dicots, monocots, cryptogams, native species and exotic species. 5 Searches for intrinsic guilds produced no evidence of guild structure, in the Pianka sense of groups of species that use similar resources, compete strongly with each other, and thus are especially subject to mutual competitive exclusion. 6 It is concluded that interspecific competition limits the biomass of each species in these grasslands, but there is no evidence from the community structure that it causes local competitive exclusion. There was no indication at either site that the communities were structured by guilds. It is concluded that, at least at the scale examined and at the time examined, Pianka‐type guilds do not exist in this vegetation.