O ne of the most difficult challenges facing ecologists is to understand and predict rapid change in Earth's ecosystems, precipitated by shifts in important drivers. Such drivers include climate; the abundance of key plant and animal species, including exotics; and atmospheric concentrations of carbon (C) and nitrogen (N), two elements that commonly limit ecosystem productivity (Vitousek et al. 1997) . Therefore, understanding the science of change is critical to predicting the future state of

more »

... cological systems and their services to society. Much of the theory and understanding in ecology, however, comes from ecosystems that are not undergoing rapid change. In these more stable ecosystems, structure and composition do not change radically over the long term, and inputs of energy and materials are approximately equal to outputs. As external drivers change, predicting the response of ecosystems in flux may be more difficult than predicting the response of stable ecosystems. Ecological studies provide a source of theory and information about ecosystems undergoing change, particularly with respect to secondary succession (Pickett and White 1985, Wallace 2004). Glacier Bay National Park and Preserve in southeastern Alaska (figure 1 ) provides a unique opportunity to study the primary successional development of newly created ecosystems. Following rapid and welldocumented glacial retreats, adjacent terrestrial, lake, stream, and intertidal habitats are exposed to biotic colonization and community development. A suite of largely independent studies conducted in these different ecosystems at Glacier Bay has shown surprisingly similar patterns of development, characterized by accumulation of organic matter, changes from pioneer species to taxa characteristic of more complex communities, and the development of biotically derived structural features. These patterns suggest a general shift from physical control to increasing biotic control as succession proceeds. For example, as plants cover recently deglaciated surfaces and drive changes in soil development, adjacent young lakes become more acidic, with higher levels of dissolved organic carbon (DOC) and lower concentrations of ions, such as calcium (Ca 2+ ), that have a mineral soil origin (Engstrom et al. 2000) . Similarly, in streams, colonization is initially driven by abiotic variables, particularly water temperature and channel stability, which give way to the increased influence of biotic controls as terrestrial vegetation stabilizes stream banks