No observational evidence for diversity enhancing productivity in Mediterranean shrublands

In a recent study, Troumbis and Memtas [Oecologia (2000) 125:101-128] demonstrated that plant diversity and productivity were correlated in a Mediterranean grassland, and presented this as observational evidence that diversity may positively influence productivity. However, the hump-backed model between biomass or productivity and diversity predicts that productivity and diversity should be positively correlated in low-productivity situations because both should be limited by the same factors. The sites used by Troumbis and Memtas are extremely unproductive (aboveground productivity of all the stands that they considered is <20 g/m(2) per year), and well within the productivity range in which productivity and diversity should be correlated as according to the hump-backed model. Evidence is presented that in their study both productivity and diversity may have been driven by the same site variables without any causative relationship existing between the two.