Ecological processes generally involve different levels of organization, starting with a single species individual, through a population of many individuals of a given species, and up to a diverse community consisting of large populations of many different species interacting among themselves and with their physical environment. Upscaling low-level attributes, such as species traits and biomass-resource feedbacks, to community level properties, such as vegetation patterns and species diversity, is a highly challenging goal for theorists. In this talk I will describe recent progress our group has made towards achieving this goal in the context of dryland plant communities.