seed density trials

1 Few experiments directly test the role of density dependence in natural plant communities.
2 We tested (i) whether different initial seed densities of the entire seed bank of an annual plant community affected performance (emergence, mean plant biomass and survival) and (ii) whether density-dependent processes were operating within the communities. We also tested whether life-stage, year and soil moisture (both on a topographic gradient and in an experimental manipulation) influenced the effects of seed density.
3 We considered two distinct phases: seed to emergent seedlings, and emergent seed- lings to established plants.
4 The seed bank was collected from a semistabilized sand dune in the Negev Desert, Israel. This was added to sieved sand in plots at the same site at four different initial seed densities: 1/16�, 1/4�, 1� (natural seed density) and 2�. The experiment was repeated for three consecutive growing seasons.
5 Emergence of seedlings was significantly influenced by initial seed density in all 3 years, with higher initial seed densities having lower rates of emergence.
6 Mean final plant size was negatively density dependent and consistently unaffected by the initial seed density sown.
7 In general, there were no strong interactions of topographic position with initial seed density, processes within the vegetation were not density dependent and the experi- mental addition of water did not influence any of the performance measures tested. Density-dependent processes vary from year to year, while moisture effects do not.
8 We conclude that seedling emergence and some processes in the established plant community are density dependent, but the established plant community is also affected by other processes, such as resource limitation.