Predicting the Vulnerability of Streams to Episodic Acidification and Potential Effects on Aquatic Biota in Shenandoah National Park, Virginia

Abstract

Acidic deposition is one of the most serious environmental problems affecting Shenandoah National Park in north-central Virginia. The park is the third most contaminated park in the National Park System because of the deposition of acid rain. Acid rain affects headwater streams in the park by temporarily reducing the acid-neutralizing capacity (ANC) of the water, a process termed episodic acidification. In turn, the increase in acidic components in streamwater can have deleterious effects on the aquatic biota.

Although acidic deposition to the park is relatively uniform across its land area, the water-quality response of streamwater during rain events varies substantially. This response is a function of the underlying geology and topographic attributes of watersheds.

Geologic and topographic data for the park's 231 watersheds are readily available; however, long-term (years and tens of years) measurements of streamwater ANC and accompanying discharge are not and would be prohibitively expensive to collect. Modeled predictions of the vulnerability of the park's streams to episodic acidification are an alternative to long-term water-quality monitoring. These predictions can aid park officials in making management decisions.

In an attempt to model the magnitude, frequency, and duration of a water-quality parameter, transfer function time series models were developed to predict hourly ANC from discharge for five watersheds in the park that have long-term records of water quality and discharge. Hourly ANC predictions over short time periods were averaged and distributions of the recurrence intervals of annual minimum ANC values were modeled for periods of 6, 24, 72, and 168 hours. The distributions were extrapolated to the rest of the watersheds in the park on the basis of watershed geology and topography. These distributions allow quantitative assessments to be made of watershed vulnerability in the park, thereby providing better information to decision makers than the qualitative assessments that can be made on the basis of geology alone.

On the basis of the models, large numbers of park streams have 6- to 168-hour (1-week) periods of low-ANC values, which may stress resident brook trout and other fish populations. The results indicate that smaller watersheds are more vulnerable to episodic acidification than larger watersheds on the same bedrock. Watersheds with similar topography and area are more vulnerable if they are underlain by less basaltic/carbonate bedrock. Additional model results indicate that substantial areas of the park are vulnerable to successive annual episodic ANC decreases in streamwater that could cause mortality of some fish species. For example, approximately 14 percent of the park watersheds are predicted to have 72-hour periods of average ANC less than 0 microequivalents per liter (µeq/L) at least once every 2 years. At this frequency, these watersheds can be expected, with a probability greater than 90 percent, to have 4 continuous years of these conditions at least once in the next 40-100 years.

First posted March 7, 2006