9727545 Robinson Spatial relationships can have strong influences on the structure and function of ecological communities, and restoration experiments can be used to explore those influences during formative phases of community development. The context of this research is the failure of natural succession to woodland on urban, derelict lands. The experimental approach is to test methods for stimulating woodland development, studying above- and belowground processes that promote or inhibit succession. In an on-going experiment on a large tract of highly disturbed land, the Fresh Kills Landfill in New York City, clusters of woody plants have been established as successional inocula. The clusters are arrayed in different patch sizes to assess the development of mutualisms, the role of herbivory, and the limitations of soil processes on establishment and rates of restoration. Funding for the current proposal will allow these investigators to continue this valuable long-term restoration experiment. Results from this research will provide explicit practical information for the restoration of degraded lands. In addition, this project will provide experimental tests of questions derived from the large body of theory on ecological succession.