COMMUNITY ECOLOGY Detecting the role of individual species for overyielding in experimental grassland communities composed of potentially dominant species Christiane Roscher Æ Jens Schumacher Æ Wolfgang W. Weisser Æ Bernhard Schmid Æ Ernst-Detlef Schulze Received: 10 July 2006 / Accepted: 13 August 2007 / Published online: 13 September 2007 Ó Springer-Verlag 2007 Abstract Several studies have shown that the contribu- tion of individual species to the positive relationship between species richness and community biomass production cannot be easily predicted from species mono- cultures. Here, we used a biodiversity experiment with a pool of nine potentially dominant grassland species to relate the species richness–productivity relationship to responses in density, size and aboveground allocation patterns of individual species. Aboveground community biomass increased strongly with the transition from monocultures to two-species mixtures but only slightly with the transition from two- to nine-species mixtures. Tripartite partitioning showed that the strong increase shown by the former was due to trait-independent com- plementarity effects, while the slight increase shown by the latter was due to dominance effects. Trait-dependent complementarity effects depended on species composition. Relative yield total (RYT) was greater than 1 (RYT [ 1) in mixtures but did not increase with species richness, which is consistent with the constant complementarity effect. The relative yield (RY) of only one species, Arrhenatherum elatius, continually increased with species richness, while those of the other species studied decreased with species richness or varied among different species compositions within richness levels. High observed/expected RYs (RYo/ RYe [ 1) of individual species were mainly due to increased module densities, whereas low observed/expec- ted RYs (RYo/RYe \ 1) were due to more pronounced decreases in module density (species with stoloniferous or creeping growth) or module size (species with clearly- defined plant individuals). The trade-off between module density and size, typical for plant populations under the law of constant final yield, was compensated among species. The positive trait-independent complementarity effect could be explained by an increase in community module density, which reached a maximum at low species richness. In contrast, the increasing dominance effect was attribut- able to the species-specific ability, in particular that of A. elatius, to increase module size, while intrinsic growth limitations led to a suppression of the remaining species in many mixtures. Keywords Complementarity effect  Constant final yield  Density  Dominance effect  Relative yield Introduction A number of experimental studies have shown that increasing plant species diversity positively affects pro- cesses measured at the ecosystem level. One example of this effect is that productivity was often found to increase Electronic supplementary material The online version of this article (doi:10.1007/s00442-007-0846-4) contains supplementary material, which is available to authorized users. C. Roscher  W. W. Weisser Institute of Ecology, University of Jena, Dornburger Strasse 159, 07743 Jena, Germany C. Roscher (&)  J. Schumacher  E.-D. Schulze Max Planck Institute for Biogeochemistry, P.O.Box 100164, 07701 Jena, Germany e-mail: croscher@bgc-jena.mpg.de J. Schumacher Institute of Stochastics, University of Jena, Ernst-Abbe-Platz 2, 07743 Jena, Germany B. Schmid Institute of Environmental Sciences, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland 123 Oecologia (2007) 154:535–549 DOI 10.1007/s00442-007-0846-4