The goal of this research is to elucidate ways in which ecological and social systems shape each other, and to understand the mechanisms by which the structure and function of natural systems affects, and in turn are affected by, human values and socioeconomic behaviors. We hypothesize that healthy biophysical systems are related to enhanced human performance and productivity and that pronounced shifts in either ecosystem integrity or in socioeconomic stability will substantially affect the other. Our approach in this first phase of the research was to examine these relationships in 18 subwatersheds within a densely populated coastal river basin in the northeastern U.S. Each subwatershed is approximately 400 ha in area and has relatively homogeneous land use. The study areas were selected to represent a continuum of development from rural to urban conditions and of environmental quality from healthy to highly degraded. The eighteen sites have been characterized from the perspective of six disciplines through state-of-the-art field methods, observational data, and surveys of watershed residents. Altogether thousands of careful analyses were conducted in hundreds of categories across the six disciplines represented. A total of 98 summary variables for each of the 18 subwatersheds were subjected to the main statistical analyses.

The hallmark of this project is that it is truly interdisciplinary. Each of the three main science areas (physical, biological, and social) is represented by two disciplines. This equal representation by hydrology, aquatic chemistry, aquatic ecology, applied/forest ecology, economics, and social ecology was designed in from the beginning; principal investigators in each of these areas worked together to design the research program from its inception.

We see this research as a long-term project that will advance in series of stages. The first phase is represented by the just-completed field investigation that used direct observation and statistical analyses to search for links between biophysical systems (here watersheds) and human social systems. The next phase will be large scale (landscape scale) experimentation in an attempt to test causality in the correlations we have identified.

The most profound conclusion of this work is that human environmental values and behaviors, socioeconomic relationships, and quality of life indices do show strong correlations with variables that characterize the physical and biological health of ecosystem. Just as importantly, while some of these relations are direct, the vast majority work through the intermediary of landscape characteristics. It appears that humans respond to their environment through features that they can directly perceive with their senses. Likewise, humans influence the environment, at the watershed scale, through the way they use the land. Importantly, these relations are independent of population density, wealth, or education among people living in the watersheds.

Of the total of 98 summary variables examined, a small subset have a strong predictive power across the biophysical, landscape, and social datasets. The others are largely redundant and add little to our ability to monitor human or ecosystem conditions. This means that a much smaller set of data can be collected in future research of this kind by us or others.

Our research design was intentionally modeled after the Hubbard Brook experiment's paired watershed approach, but with the addition of humans as important environmental actors. This method assumes that hydrologically defined watersheds approximate self-contained systems that can be investigated through measurement of inputs and outputs. One difficulty that has been observed at Hubbard Brook and elsewhere is that some important biophysical components do not conform to watershed boundaries, and diffuse inputs and outputs (e.g., bird migrations, gas exchange) are difficult to measure directly. This difficulty can be especially acute for people, whose actions take place at multiple locations and at a variety of scales from local to global. Thus, "peoplesheds" and watersheds do not correspond in many ways. The current research suggests that despite this broader human scale of experience, people respond to and influence their environments at the scale of their resident watersheds. Thus, watershed-based research can be used to investigate human-environment interactions.

A final contribution of this project has been establishment of a deep baseline of information for our study watersheds. Our plan is continued use of these systems in future phases of research on how people interact with their environment. In this sense we view the greater New Haven Watershed as an "urban Hubbard Brook" where we will conduct research in parallel with conventional Long Term Ecosystem Research urban sites in Baltimore and Phoenix.