It may seem that a cyberinfrastructure supporting ecosystems studies will need to be built somewhat differently from one supporting the study of individual organisms. But not necessarily. iPlant already deals in relationships among thousands of organisms as well as thousands of genes in its true of life and genotype-phenotype projects, and it also has developed an innovative taxonomic resolution tool used by biodiversity scientists to resolve conflicts in taxonomic databases. Thus, it already deals with a multitude of complex interaction, and may be well positioned to take on higher levels of complexity than might at first be apparent.
Most likely, however, I suspect we will see a coalescence of projects - including the iPlant Collaborative, the National Center for Ecological Analysis and Study(NCEAS), the National Evolutionary Synthesis Center (NESCent), the National Ecological Observatory Network (NEON), the National Institute for Mathematical and Biological Synthesis (NIMBioS) and others - leading eventually to a higher order entity that comprises a large part of what iBiosphere is envisioned to be. However, iBiosphere also needs to be much more - it must incorporate the social sciences, the agricultural sciences, and the environmental sciences as equal partners in order to inform and facilitate the kinds of policy and management decisions that will permit humankind to create and maintain sustainable systems from agriculture to ecology to the biosphere.
Bringing together all these existing, developing entities in a meaningful and functional manner to allow complex decision making may seem like a nearly impossible task. But importantly, it is unlikely to happen by 'top down' planning; rather, it is more likely to just 'come together', through many efforts of many groups of people, in ways that are difficult or impossible to envision today. We can facilitate this coming together, but my guess is that we shouldn't try too hard to predict how exactly it is going to happen, or exactly what it is going to look like in the end. Taking a page from iPlant's original playbook, we should let the community decide what is needed, what the challenges are, and what the priorities ought to be, and then we should facilitate those efforts by responding to needs as they arise, not by deciding in advance what those needs might be.