The species richness–biomass relationship in herbaceous plant communities: what difference does the incorporation of root biomass data make?

So far, in all studies on the much-discussed hump-backed relationship between plant community productivity and species richness, productivity has been assessed through plant shoot biomass, i.e. it has been ignored that frequently most of the biomass is produced below ground. We revisited the 27 grassland and forest field-layer communities, studied earlier by Zobel and Liira, to sample root biomass, plant total biomass and root/shoot allocation, and learn how the incorporation of below-ground biomass data would affect the shape of the hump-backed relationship. In order to avoid scaling artefacts we estimated richness as the average count of species per 500 plant ramets (absolute richness). We also included relative richness measures. Relative richness was defined as richness per 500 ramets/size of the actual species pool (the set of species present in the community), relative pool size was defined as size of the actual species pool/size of the regional species pool (the set of species available in the region and capable of growing in the given community).
The biomass-absolute richness relationship was humped, irrespective of the biomass measure used, the hump being most obvious when plant total biomass was used as the independent variable. Evidently, the unimodal richness–productivity curve is not a sampling artefact, as suspected by Oksanen. However, relative richness was not related to community biomass (above-ground, below-ground or total). The hump-backed curve is shaped by the sizes of actual species pools in communities, implying that processes which are responsible for small-scale diversity pattern mainly operate on the community level.
Neither absolute nor relative richness were significantly related to root/shoot allocation. The presumably stronger (asymmetric) shoot competition at greater allocation to shoots appears not to suppress small-scale richness. However, there is a significant relationship between relative pool size and root/shoot allocation. Relatively more species from regional species pools are able to enter and persist in communities with more biomass allocated into roots.