The relation between community biomass and evenness: what does community theory predict, and can these predictions be tested?

Several theories of community structure make predictions of how the processes that control species diversity should change along a gradient of decreasing stress and therefore, apart from disturbance effects, along a gradient of increasing biomass. These include the theories of 'Small species pool', 'Higher extinction rate', 'Small populations', 'Reduced patchiness and 'Competitive dominance'. This body of theory has previously been applied to species richness, giving the Humped-back model. It has not previously been applied to evenness. We do this, and show that the same theory predicts a monotonic decrease in evenness as stress decreases and biomass increases. We test this prediction, by sampling 10 distinct vegetation types in a temperate oceanic area of New Zealand, selected to cover the widest range of ecological conditions in the region. In replicate quadrats, the photosynthetic biomass of all species was determined. Evenness was calculated by three recently recommended indices. Evenness was lower in sites with higher biomass, in superficial conformity with theory, However, a permutation test showed that this relation persisted when the biomass values were allocated to sites and species at random. This artefact is shown to be related to the approximately geometric distribution of abundances in most plant communities. There is no tendency in the observed data for evenness to decrease more strongly than expected under this null model. Our results therefore fail to support the 'Higher extinction rate', 'Small populations', 'Reduced patchiness' and 'Cumulative competition' theories on which the prediction was based.