The U.S. Forest Service is concerned about the current and future health of terrestrial and aquatic resources within the Blue Ridge Province of the Southern Appalachian Mountains in western North Carolina, eastern Tennessee, and South Carolina. Soils within some of these watersheds are inherently low in base cations. Adequate amounts of available calcium, magnesium, and potassium are essential to maintain healthy vegetation, and calcium is sometimes a factor limiting the development of healthy aquatic organisms. There is concern that the rate of base cation loss from the soil in the more acid-sensitive watersheds has been accelerated by the deposition of sulfur (S) and nitrogen (N) compounds from the atmosphere, and that some soils may be experiencing base cation depletion.
One way to determine the ability of a watershed to buffer acid inputs and the adequacy of the soil base cation supply is to measure the acid-neutralizing capacity (ANC) of the drainage water. Streams having ANC values greater than 50 microequivalents per liter (µeq/L) are considered to have adequate buffering capacity to offset the deposition of S and N compounds, but some highly sensitive aquatic organisms may be adversely impacted at ANC values near or below 50 µeq/L. Low streamwater ANC can signal the possibility of base-poor watershed soils. There are existing data on many hundreds of streams within the region, and these data provide the foundation for both aquatic and terrestrial resource evaluation.
An important tool in the evaluation of acidification damage to aquatic and terrestrial ecosystems is the critical load, which can be defined as the level of acidic deposition below which ecological damage would not be expected to occur. The critical load for protection of aquatic biota is generally based on maintaining surface water ANC at an acceptable level.
This E&S study included calibration and application of the watershed model, MAGIC, to estimate the sensitivity of 66 watersheds in the Blue Ridge Province to changes in atmospheric S deposition. The principal objectives of this research report were to:
MAGIC model simulations estimated that stream ANC values were above 20 µeq/L in all modeled watersheds in 1860, but below 50 µeq/L in 38% of the watersheds and below 100 µeq/L in 86% of the watersheds at that time. The minimum simulated ANC in 1860 among the modeled streams was 30 µeq/L. These hindcast water chemistry results are important, not only to assess the extent to which resources have been damaged by air pollution, but also to provide constraints regarding expectations for future chemical recovery. The hindcast simulation results suggested that the average of the modeled streams was acidified from ANC=65 µeq/L in 1860 to ANC=36 µeq/L in 2005.
In general, the model projected that soil and stream chemistry have changed substantially since pre-industrial times, but that future changes in response to emissions controls will be small. Simulation results suggested that modeled watersheds would not change to a large degree with respect to stream ANC or soil % base saturation, depending on the extent to which emissions are reduced in response to three emissions control scenarios (Base Case, Moderate, and Aggressive Additional Controls).
Site-to-site variability in critical loads was very high. Estimated critical loads for S deposition ranged from less than 0 kg S/ha/yr (ecological objective not attainable) to more than 1,000 kg S/ha/yr, depending on the selected site, ANC endpoint, and evaluation year. Thus, for some sites, one or more of the selected target ANC critical levels (0, 20, 50, 100 µeq/L) could not be achieved by the year 2100 (or alternative evaluation year) even if S deposition was reduced to zero and maintained at that level throughout the simulation. For other sites, the watershed soils contained sufficiently large buffering capacity that even very high sustained levels of atmospheric S deposition would not reduce stream ANC below common damage threshold criteria values.
In order to aid in the process of extrapolating MAGIC model critical loads simulation results to watersheds within the study area that were not modeled, we developed a suite of multiple regression equations to estimate critical loads from variables that are more widely available across the region than are the MAGIC model results. Separate multiple regression prediction efforts were conducted to estimate critical load from 1) landscape variables represented spatially in the GIS, 2) a combination of landscape variables and stream chemistry variables, and 3) streamwater chemistry only. Based on these relationships, critical loads were estimated for the broader region.