 |
 |
 |
 |
 |
 |
 |
Wild Trout VII, Yellowstone National Park, October 2000
J.R. Webb, F.A. Deviney, Jr., B.J. Cosby, A.J. Bulger, J.N .Galloway
Department
of Environmental Sciences
University of Virginia
Charlottesville, Virginia 22903
Regionalization
Stoddard et al. [1999] proposed two criteria that should be fulfilled before site-specific trends from multiple sites are combined into single estimates of regional trends. First, the sites must represent the region or regional subpopulation of interest. Second, trend behavior should be consistent among sites.
We have examined the possibility of obtaining single estimates of trends for three regional subpopulations that we defined based on watershed lithology. These broad classes (siliciclastic, crystalline, and minor-carbonate) include the predominant bedrock in all quarterly sampled watersheds and encompass most of the brook trout range in western Virginia. As depicted below, ANC distributions differ for streams associated with these three classes. We thus expect the response to acidic deposition to differ among the classes.
Following the methods described by Stoddard et al. [1999], we have tested sulfate and ANC trends for consistency among sites and among seasons within the subgroups.
Within the three subgroups, trends in sulfate are neither consistent by site nor by season. ANC trends, however, are consistent by site, but not by season. When we examined ANC trends by season, we found a significant negative trend (p<0.05) for winter in all three groups. Thus, the median ANC trend slopes provide estimates of regional winter-season trends. For the subpopulations of streams associated with siliciclastic, crystalline, and minor-carbonate lithologies, these trends in ANC are –0.65, -1.01, and –0.91 µeq/L/yr, respectively.
It is notable that these trends occur during the winter season when early, acid-sensitive life-stages of the brook trout are present.
|
