The Effect of Landscape Features on Deposition to Hunter Mountain, Catskill Mountains, New York

Abstract

Atmospheric deposition to montane ecosystems is higher than to adjacent lowlands. However, because of the heterogeneity of mountainous landscapes, rates of deposition are likely to vary considerably across major landscape features. Estimates of total atmospheric deposition for montane areas in the United States are wide ranging and usually based on models that do not take into account landscape heterogeneity. Thus, little is known about the spatial variability of atmospheric deposition to these high-elevation ecosystems. We identified four landscape features that are likely to control patterns and rates of atmospheric deposition in mountainous terrain, including (1) forest edges or gaps, (2) elevation, (3) aspect, and (4) vegetation type, and we measured patterns of atmospheric deposition across them in the Catskill Mountains of New York State. We measured lead amount in the forest floor as an index of atmospheric deposition, used these data to estimate relative deposition to the Hunter Mountain massif, and with a geographic information system (GIS) database, scaled up atmospheric deposition estimates to the Hunter Mountain landscape. Area-weighted mean deposition to the region above 1000 m (high-elevation) was 13–43% greater than to a nearby, low-elevation site, depending upon how edges were defined. In addition, we identified “hotspots” of deposition, such as high-elevation, coniferous forest edge zones, where atmospheric deposition of pollutants and nutrients was 300% greater (44 kg N·ha−1·yr−1 and 46 kg S·ha−1·yr−1) than to low-elevation forests (11.1 kg N·ha−1·yr−1 and 11.6 kg S·ha−1·yr−1). Our data suggest that estimates of atmospheric deposition to mountainous terrain that do not include landscape heterogeneity may be inadequate. Furthermore, current models of atmospheric deposition cannot predict “hotspot” regions of extremely high deposition. Landscape features such as forest edges, elevation, aspect, and vegetation type should be considered in future modeling efforts.