23: Nitrogen Input-output Budgets for Forests in the Chesapeake Bay Watershed

We analyzed three kinds of information to evaluate the potential of forests in the Chesapeake Bay watershed to retain the nitrogen in atmospheric deposition. Long-term (13:~) measurements of nitrogen in precipitation and stream discharge for one forested watershed in the Inner Coastal Plain showed consistently high nitrogen retention. Annual wet precipitation deposited an averagk of U.4 kg N/ha, of which 46% was nitrate and 26% ammonium. Dry deposition of nitrate estimated by net throughfall was 24% of wet nitrate deposition. The forest lost an average of 1.95 kg N/ha in annual stream discharge, of which 83% was organicN. Stream discharges of nitrate and ammonium were only 3.2 and 5.4% of wet deposition inputs. A 1-year broad-scale survey indicated low nitrate outputs from forested watersheds in a variety of geological settings within the Chesapeake drainage. Nitrate concentrations were markedly higher in streams draining watersheds with some non-forested land, particularlv cultivated lands. Consistentlv low nitrate o u t ~ u t s from a variety of forested watersheds suggest thHt high ni&ogen retention is a common feature of forests k the Chesapeake ~a~ watershed. A literature study indicated that of 25 forested watersheds where precipitation inputs and stream discharges were measured, 21 retained most of the inorganic N from atrnosp'heric' inputs. * A worldwide concern has developed over how forests will respond to long-term high rates of atmospheric deposition, particularly of nitrogen and acid. According to a paradigm proposed in the 1980s, young aggrading forests efficiently retain nutrient inputs, but as forests mature they retain less and less of any nutrient inputs 111. More recently, a hypothesis has developed that forests become nitrogen saturated as a result of chronic high rates of atmospheric deposition [2-61. A nitrogen saturated forest stops exhibiting net incorporation of nitrogen inputs, so the flux of nitrogen in watershed discharge approximately equals the nitrogen input from atmospheric deposition. However, deposited nitrogen may still be assimilated and recycled before being discharged [7, 81. A forest's response to chronically high atmospheric loading also depends on the tree species composition. Much of what is known about nitrogen cycling within various types of forests has been summarized in recent reviews [g-U]. One way to evaluate the effects of atmospheric deposition on forest ecosystems is to measure input and output mass balances, thereby treating the forest as a black box. This mass balance approach is best applied at a watershed level by comparing atmospheric inputs with watershed stream discharges. One of the best examples of an input-output analysis of a forest ecosystem was Hubbard Brook in the White Mountains of New Hampshire, where bulk precipitation inputs and accurately gauged stream outputs were measured for U years [U] . Many other input-output studies of forested watersheds have been done for shorter periods or with less thorough sampling. For example, one 3year study in Sweden compared bulk precipitation to stream outputs, which were measured by continuous stream gauging (but only biweekly grab sampling of stream chemistry (131.) Another Swedish study of three gauged basins sampled bulk precipitation and stream chemistry only monthly for 10 years [14]. In the White Mountains of New Hampshire, bulk ~recipitation was measured, but stream discharge volumes were extrapolated from nearby Hubbard Brook, and stream chemistry was grab sampled biweekly [15]. A %year study in the Vosges Massif in north-eastern France measured bulk precipitation inputs and stream outputs, but fluxes of ammonium and nitrate must be inferred from other data in the article [16]. Balances were measured for two years in Massachusetts 1171, for three basins in Sweden [18], and for three years in the Colorado Rockies [19]. All of these studies reported data for both nitrate and ammonium, but most did not report organic-N. Clearly, some were completely forested watersheds, but for others, land use composition was unclear.