Vegetation dynamics associated with changes in atmospheric nitrogen deposition and climate in hardwood forests of Shenandoah and Great Smoky Mountains National Parks, USA.

Ecological effects of atmospheric nitrogen (N) and sulfur (S) deposition on two hardwood forest sites in the eastern United States were simulated in the context of a changing climate using the dynamic coupled biogeochemical/ecological model chain ForSAFE-Veg. The sites are a mixed oak forest in Shenandoah National Park, Virginia (Piney River) and a mixed oak-sugar maple forest in Great Smoky Mountains National Park, Tennessee (Cosby Creek). The sites have received relatively high levels of both S and N deposition and the climate has warmed over the past half century or longer. The model was used to evaluate the composition of the understory plant communities, the alignment between plant species niche preferences and ambient conditions, and estimate changes in relative species abundances as reflected by plant cover under various scenarios of future atmospheric N and S deposition and climate change. The main driver of ecological effects was soil solution N concentration. Results of this research suggested that future climate change might compromise the capacity for the forests to sustain habitat suitability. However, vegetation results should be considered preliminary until further model validation can be performed. With expected future climate change, preliminary estimates suggest that sustained future N deposition above 7.4 and 5.0 kg N/ha/yr is expected to decrease contemporary habitat suitability for indicator plant species located at Piney River and Cosby Creek, respectively.

Spatial variation of atmospheric nitrogen deposition and critical loads for aquatic ecosystems in the Greater Yellowstone Area.

Current and historic atmospheric nitrogen (N) deposition has impacted aquatic ecosystems in the Greater Yellowstone Area (GYA). Understanding the spatial variation in total atmospheric deposition (wet + dry) of N is needed to estimate air pollution deposition critical loads for sensitive aquatic ecosystems. This is particularly important for areas that have an increasing contribution of ammonia dry deposition to total N (TN), such as the GYA. High resolution geostatistical models and maps of TN deposition (wet + dry) were developed using a variety of techniques including ordinary kriging in a geographic information system, to evaluate spatial variability and identify areas of elevated loading of pollutants for the GYA. TN deposition estimates in the GYA range from <1.4 to 7.5 kg N ha-1 yr-1 and show greater variability than wet inorganic N deposition. Critical loads of TN deposition (CLTNdep) for nutrient enrichment in aquatic ecosystems range from less than 1.5 ± 1.0 kg N ha-1 yr-1 to over 4.0 ± 1.0 kg N ha-1 yr-1 and variability is controlled by differences in basin characteristics. The lowest CLTNdep estimates occurred in high elevation basins within GYA Wilderness boundaries. TN deposition maps were used to identify critical load exceedances for aquatic ecosystems. Estimated CLTNdep exceedances for the GYA range from 17% to 48% depending on the surface water nitrate (NO3-) threshold. Based on a NO3- threshold of 1.0 µmol L-1, TN deposition exceeds CLTNdep in approximately 30% of the GYA. These predictive models and maps can be used to help identify and protect sensitive ecosystems that may be impacted by excess atmospheric N deposition.

Assessment of lake sensitivity to acidic deposition in national parks of the Rocky Mountains.

The sensitivity of high-elevation lakes to acidic deposition was evaluated in five national parks of the Rocky Mountains based on statistical relations between lake acid-neutralizing capacity concentrations and basin characteristics. Acid-neutralizing capacity (ANC) of 151 lakes sampled during synoptic surveys and basin-characteristic information derived from geographic information system (GIS) data sets were used to calibrate the statistical models. The explanatory basin variables that were considered included topographic parameters, bedrock type, and vegetation type. A logistic regression model was developed, and modeling results were cross-validated through lake sampling during fall 2004 at 58 lakes. The model was applied to lake basins greater than 1 ha in area in Glacier National Park (n = 244 lakes), Grand Teton National Park (n = 106 lakes), Great Sand Dunes National Park and Preserve (n = 11 lakes), Rocky Mountain National Park (n = 114 lakes), and Yellowstone National Park (n = 294 lakes). Lakes that had a high probability of having an ANC concentration <100 microeq/L, and therefore sensitive to acidic deposition, are located in basins with elevations >3000 m, with <30% of the catchment having northeast aspect and with >80% of the catchment bedrock having low buffering capacity. The modeling results indicate that the most sensitive lakes are located in Rocky Mountain National Park and Grand Teton National Park. This technique for evaluating the lake sensitivity to acidic deposition is useful for designing long-term monitoring plans and is potentially transferable to other remote mountain areas of the United States and the world.

A screening procedure for identifying acid-sensitive lakes from catchment characteristics.

Monitoring of Wilderness lakes for potential acidification requires information on lake sensitivity to acidification. Catchment properties can be used to estimate the acid neutralizing capacity (ANC) of lakes. Conceptual and general linear models were developed to predict the ANC of lakes in high-elevation (> or = 2170 m) Wilderness Areas in California's Sierra Nevada mountains. Catchment-to-lake area ratio, lake perimeter-to-area ratio, bedrock lithology, vegetation cover, and lake headwater location are significant variables explaining ANC. The general linear models were validated against independently collected water chemistry data and were used as part of a first stage screen to identify Wilderness lakes with low ANC. Expanded monitoring of atmospheric deposition is essential for improving the predictability of lake ANC.