Effectiveness and Longevity of Phosphorus Inactivation with Alum

Abstract

ABSTRACT Effectiveness and longevity of alum treatments were evaluated in 21 lakes (or lake basins) across the United States; 9 were polymictic and 12 were dimictic. Effectiveness was judged from reductions in lake TP (total phosphorus) and internal loading rate, as well as chlorophyll a (chl a), both initially and over periods ranging from 4 to 20 years following treatment. Internal loading rate was reduced in six of nine polymictic lakes/basins by an average of two-thirds, and lake TP was reduced by about one-half, which persisted for 5–11 years. Internal loading rate in dimictic lakes (7 of 7 with adequate data) remained reduced by an average of 80% for 4 to 21 years (average 13 yrs). For the six polymictic lakes, in which treatment was effective, chl a decreased by an average of two-thirds initially, but was about 40% less than die pre-treatment level after 5 to 11 years. Chl a decreased in seven dimictic lakes by an average of 57% initially and 42% after 5–18 years. In some cases, response was independent of alum treatment because the trophogenic zone was affected more by external than internal P loading. At West Twin Lake, OH, nutrient diversion was more important than control of P release from hypolimnetic sediments in improving surface water quality. The significance of vertical P entrainment, as well as external loading, should be considered before treating dimictic lakes. Alum was initially ineffective in two polymictic lakes and was effective only for a few years in a third, apparently due to interference from macrophytes. Dense macrophytes may have caused uneven floe distribution or sediment P was recycled from below the floe layer through plant senescence and decay. Where macrophytes did not interfere, treatments were usually effective for at least 8 years and up to 11 years in polymictic lakes; an expectation of 10 years longevity is considered reasonable for unstratified lakes. Several dimictic lakes showed effectiveness/longevity of internal loading control for at least 13 and up to 18 to 20 years; an expectation of 15 years longevity seems reasonable for dimictic lakes. The decline in treatment effectiveness could have been due to bioturbation, or to sinking and burial of the aluminum floe over time, as evidenced by a distinct Al layer located at about 20 cm after 16 years in West Twin Lake, OH. A direct relation between dose and treatment longevity, as well as an inverse relation between dose and sediment P release rate, was apparent in the dimictic lakes.