A 10 000‐Yr History of Natural Ecosystem Acidification

Abstract

This study examines natural, long—term (≈104 yr) acidification processes in New England using paleoecological methods (pollen, diatoms, and the chemistry of three operationally defined sediment fractions) and a paired watershed approach.The primary variable is lithology, and the focus of this study is to illuminate the role of lithology in long—term ecosystem acidification. Unlike most cases of recent (anthropogenic) acidification in which inferred processes of ecosystem acidification occur over short time spans (10—50 yr), natural acidification processes develop over hundreds or thousands of years, providing enough temporal resolution in the stratigraphic record to allow strong inferences about the factors controlling the occurrence, timing, and magnitude of terrestrial and aquatic ecosystem acidification. The two sites in this study are similar in size, lake depth, elevation, aspect, and local climate, but Cone Pond, New Hampshire, is an acidic, clearwater lake in a catchment of thin tills derived from base—poor gneisses and schists, whereas South King Pond, Vermont is a mesotrophic lake in a catchment of thicker tills derived from slates, phyllites, and limestone. Although vegetation history is generally similar between the two sites, the Cone Pond catchment included higher representation of acidophilic vegetation than the South King Pond catchment throughout its 10 000—yr history. There are no known exposures of sulfide minerals in the Cone Pond watershed, nor are the basal inorganic sediments S—rich. Both diatom and geochemical analyses indicated significant long—term acidification of the upland soils and surface waters of the Cone Pond, but not the South King Pond, catchment. At Cone Pond, the uplands exported A1 almost entirely in labile form by 10 000 yr before present (BP). Increases in labile Al 8000—5500 BP, followed by substantial declines in sedimentary MN and in the ratio of Ca to organic matter, indicate soil acidification followed by decreasing lakewater pH. The loss of the Melosira/Cyclotella diatom assemblage ≈7200 BP as hemlock populations began to expand, coupled with the disappearance of the chrysophyte Mallomonas torquata, suggests increasing lakewater transparency possibly accompanied by decreasing lakewater pH. By 5000 BP, all planktonic diatom species has disappeared, and the acidobiontic diatom flora was well established. Diatom—inferred pH reconstructions indicate that the pH of Cone Pond was <5.5 by 6500 BP and may have dropped below 5.0 as long ago as 2000 BP, when spruce recolonized the catchment. A sudden decline in the ratio of acid—labile nonbiogenic Si to acid—labile A1 also occurred ≈2000 BP, and appears to reflect dissolution of an amorphous aluminosilicate complex, providing independent evidence for a precultural pH of ≈5.0. None of these features are found at South King Pond, where the only paleolimnological evidence of acidification is a sharp decline in sedimentary carbonates ≈5000 BP. This decrease, however, appears to be related to climatic factors rather than to lakewater pH. The historical accident of low till deposition in the Cone Pond catchment has been an important predisposing factor for natural, long—term ecosystem acidification, leading to truncated soil profiles, shortened hydrologic flow paths, lower overall supplies of base cations, and an enhanced susceptibility to biogeochemical changes driven by changes in upland vegetation. The potential role of other factors (depletion of base cations, sulfide minerals in the catchment of the aquifer feeding the ephemeral inlet, weathering of peat deposits in the upper reaches of the inlet, and historic A1 acidity) is also discussed.