PARENT SESSION
Oral Session #76: Ecosystem Ecology: Larger scale processes, geomorphology, soils.
Presiding: W. Straw
Thursday, August 8. 9:00 AM to 11:30 AM. Gila Meeting Room, TCC.
Scaling-up point measurements of atmospheric deposition to mountainous landscapes: an empirical modeling approach.
SIMKIN, SAMUEL*,1, WEATHERS, KATHLEEN1, LOVETT, GARY1, LINDBERG, STEVEN2, LEWIS, DAVID3, SCHWARZ, KIRSTEN1, 1 Institute of Ecosystem Studies, Millbrook, NY2 Oak Ridge National Laboratory, Oak Ridge, TN3 University of Massachusetts, Amherst, MA
ABSTRACT- Refining ecosystem budgets and evaluating biological responses to chronic atmospheric deposition require accurate deposition values. Total atmospheric deposition of nutrients and pollutants is highly heterogeneous, particularly in mountainous areas with complex terrain and heterogeneous vegetation. Although wet deposition has been monitored extensively since the 1980s, measurements of the more variable cloud and dry components of total deposition are sparse. We developed an empirical modeling approach that characterizes total deposition as a function of landscape features such as vegetation type, elevation, topographic exposure, slope, and aspect. Specifically, we used sulfate fluxes in throughfall and lead in surface soils as indices of total deposition to the landscapes of Acadia (ACAD) and Great Smoky Mt. (GRSM) National Parks, respectively. We made 300+ point measurements of the chosen deposition index and corresponding landscape attributes at each park, and then constructed a general linear model relating the deposition index to the landscape variables. At both parks the deposition indices ranged over an order of magnitude, with deposition varying significantly in response to vegetation type, elevation, and topographic exposure, but showing essentially no response to slope or aspect. The general linear model further indicated that elevation and vegetation type are the landscape variables most strongly influencing deposition, together explaining >45% and >35% of the variation in deposition at GRSM and ACAD, respectively. At ACAD, >70% of the variance in deposition could be explained when only samples from tall forests (> 20 m) were retained in the model. We used a GIS database of elevation and vegetation type to scale-up point measurements of total deposition fluxes and create a parkwide map of total deposition for each park.
KEY WORDS: atmospheric deposition, landscape, elevation, vegetation