Learning to roll with the punches: adaptive experimentation in human-dominated systems

The interdisciplinary study of human–environment interactions is becoming increasingly important around the world. Long-term experimental manipulations that combine approaches from both the ecological and social sciences could play an important role in the study of human–environment feedbacks in cities. The inclusion of in situ human subjects in this research is vital, as it facilitates more accurate scientific models and might produce social benefits such as increasing public trust in scientists. Within a landscape experiment, human subjects may alter experimental conditions to suit their needs, imitating the rapidly changing environmental conditions in cities. In response, researchers adjust explanatory models in a process which could be called “adaptive experimentation”. These ideas are illustrated by a description of a proposed experiment in the Phoenix metropolitan area, where residential landscaping will be manipulated and the feedbacks between ecological processes and the activities of resident humans studied.