“…Past concern over the effects of atmospheric deposition has resulted in the development of several process-based models that predict effects of various deposition scenarios on water quality (Christophersen and Wright, 1981;Cosby and others, 1985;Gherini and others, 1985). The structure of these models is typically derived from mechanisms studied in plo* experiments and extrapolated to the small-watershed scale through "lumping" or statistical averaging, but the effects of acid precipitation in many small watersheds, including those sampled in the Catskills, show a spatial variation that is not random, but rather is systematically related to stream position on a hillslope (Johnson and others, 1981;Lawrence and others, 1986; U.S. Geological Survey, unpublished data); this variation makes spatial averaging inappropriate.…”