- Biodiversity and ecosystem functioning: It is time for dispersal experiments - 543 FORUM Journal of Vegetation Science 17: 543-547, 2006 © IAVS; Opulus Press Uppsala. FORUM Biodiversity and ecosystem functioning: It is time for dispersal experiments Zobel, Martin * ; Öpik, Maarja 1,2 ; Moora, Mari 3 & Pärtel, Meelis 4 Institute of Botany and Ecology, University of Tartu, 40 Lai St., Tartu 51005, Estonia; 2 Present address: Scottish Crop Research Institute, Errol Road, Invergowrie, DD2 5DA Dundee, UK 1 E-mail maarja.opik@ut.ee; 3 E-mail mari.moora@ut.ee; 4 E-mail meelis.partel@ut.ee * Corresponding author; E-mail martin.zobel@ut.ee Abstract The experimental study of the relationship between bio- diversity and ecosystem function has mainly addressed the effect of species and number of functional groups. In theory, this approach has mainly focused on how extinction affects function, whereas dispersal limitation of ecosystem function has been rarely discussed. A handful of seed introduction experiments, as well as numerous observations of the effects of long-distance dispersal of alien species, indicate that eco- system function may be strongly determined by dispersal limitation at the local, regional and/or global scales. We suggest that it is time to replace biodiversity manipu- lation experiments, based on random draw of species, with those addressing realistic scenarios of either extinction or dispersal. Experiments disentangling the dispersal limitation of ecosystem function should have to take into account the probability of arrival. The latter is defined as the probability that a propagule of a particular species will arrive at a particu- lar community. Arrival probability depends on the dispersal ability and the number of propagules of a species, the distance a species needs to travel, and the permeability of the matrix landscape. Current databases, in particular those in northwest- ern and central Europe now enable robust estimation of arrival probability in plant communities. We suggest a general hypothesis claiming that dispersal limitation according to arrival probability will have ecosys- tem-level effects different from those arising due to random arrival. This hypothesis may be rendered more region-, land- scape- or ecosystem-specific by estimating arrival probabili- ties for different background conditions. Keywords: Arrival probability; Ecosystem function; Diaspore; Invasion; Propagule; Seed; Species diversity; Species pool. Introduction There is an increasing consensus that biodiversity significantly influences ecosystem function (Hooper et al. 2005; Srivastava & Vellend 2005), including stocks of energy and materials (e.g. biomass), fluxes of energy or materials (e.g. productivity, decomposition), and the stability of rates or stocks over time (Pacala & Kinzig 2002). The relationship between biodiversity and eco- system function implies that a reduction in biological diversity (variety of species, genotypes, etc.) will cause a reduction in ecosystem-level processes. Synthetic ex- perimental communities, exhibiting different levels of species diversity and functional group diversity, have been the most widely used object to investigate the relationship between biodiversity and ecosystem func- tion. In very general terms, research on biodiversity- ecosystem function has so far focused mainly on how extinction affects ecosystem function (Loreau et al. 2001; Schmid & Hector 2004; Srivastava & Vellend 2005). Although the role of arrival sequence (Grime 1998) or recruitment limitation in gaps (Symstad & Tilman 2001) in determining ecosystem function has been discussed, and the arrival sequence has been shown to influence species composition (Ejrnæs et al. 2006), the setup of experiments with synthetic communities has not incorporated natural processes like dispersal limitation. At the same time, we know that the species compo- sition and diversity of natural plant communities may be significantly dispersal-limited (Tilman 1997; Stampfli & Zeiter 1999; Zobel et al. 2000; Foster 2001; Foster & Dickson 2004; Lord & Lee 2001; Xiong et al. 2003). Since species functional characteristics strongly influ- ence ecosystem function (Grime 2001; Garnier et al. 2004; Lavorel & Garnier 2002; Diaz et al. 2004), the differential arrival of species may influence ecosystem properties even if the arrival of species is a random