Randomized intervention analysis of the response of the West Bear Brook Watershed, Maine to chemical manipulation

A paired watershed manipulation study was conducted to study the qualitative and quantitative impacts of elevated acidic precipitation on the chemistry of soils, soil water, and stream water. The Bear Brook Watershed, Maine (BBWM) is drained by two first order streams, East Bear Brook and West Bear Brook. The streams were chemically and hydrologically monitored for two years (1987-1989) and exhibited similar behavior. The West Bear watershed was then chemically manipulated with the bimonthly addition of (NH₄)₂SO₄ (150 mol ha^-1 per application). To assess whether changes in stream water chemistry occurred following the chemical manipulation, and if so when, Randomized Intervention Analysis (RIA) was performed using time-paired data from the two watersheds. RIA, along with autocorrelation analysis, statistically evaluates the behavior of the various analytes under the influence of artificial acidification and therefore provided an objective basis for determining whether changes in the geochemical behavior of West Bear Brook were temporally associated with the chemical manipulation. RIA analysis using weekly data yielded higher probabilities of stream water chemistry effects being temporally linked with the manipulation than RIA analysis using monthly data. Using monthly data, there is a lower probability that short-term excursions in water chemistry related to hydrology can be detected. According to RIA analysis of weekly data for three years of manipulation (1989-1992), the statistically-determined order of impact on water chemistry was (K⁺, Mg²⁺, Na⁺, Ca²⁺, total Al, pH, SO₄^2-, NO₃^-, DOC)>Si>Cl^-. Autocorrelation analysis indicated that several analytes exhibited increasingly deterministic behavior, including SO₄^2-, base cations, and DOC. Both RIA and autocorrelation analysis indicated no temporal relationship between the manipulation and hydrology.