Acid Deposition: Long-Term Trends (1986)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

9 Paleolimnological Evidence for Trends in Atmospheric Deposition of Acids and Metals Donald F. Charles and Stephen A. Norton INTRODUCTION Paleolimnological analyses of lake sediments have traditionally been used to reconstruct many aspects of the evolution of lake/watershed ecosystems, including terrestrial and aquatic vegetational succession (Davis et al. 1975), fire history (Patterson 1977), trophic status (Davis and Norton 1978, Stockner and Benson 1967), lake acidification (Battarbee 1984), and even the occurrence of blight or disease (Bradstreet and Davis 1975). Long- term changes in meteorology, morphology of the lake basins, soil characteristics, land use, and surface water chemistry can be partially determined from the sediment record. The sediments of a lake contain information on the lake's past: its biota, water chemistry, watershed characteristics, and material deposited directly from the atmosphere (Frey 1969, Pennington 1981). The information is provided by the organic and inorganic substances, in dissolved or particulate form, that entered or were formed within a lake and were deposited in its sediments. The primary materials are controlled by watershed geology, climate, and biological processes. Once deposited at the bottom of lakes, sediments can be affected by secondary *The introduction and the section on comparison of diatom and chemical data were jointly authored by Donald F. Charles and Stephen A. Norton. D. Charles authored the section on diatoms and chrysophytes. S. Norton prepared the sections on chemical stratigraphy of lake sediments and peat bogs. 335

336 processes, such as transport (horizontal and vertical) and a variety of chemical and biological activities. Sediments can be sampled by taking cores; the core samples represent the time history of deposition of sediments, with the older sediments lying at greater depth. The times when particular intervals of sediment were deposited can be determined either from analysis of radioactive decay products (lead-210, cesium-137) or by using changes in sediment characteristics that correspond to well-dated local events, such as pollen and charcoal as indicators of logging and forest fires. Nearly all lakes in the Northern Hemisphere, except those in karst topography, were formed by Pleistocene glaciation. Thus, in most lakes in the North Temperate Zone, the stratigraphic record represents the period from formation of the lake to the present (spanning 10,000 to 15,000 years or less). The period of concern in terms of recent anthropogenic acidification is about 50 to 200 years. In lakes with typical sedimentation rates this smaller interval is usually represented within the top one-half meter of sediment, with the time resolution possible in sediment studies depending on the sedimenta- tion rate, extent of mixing, subsampling interval, and type and quality of the dating of the sediment. Resolu- tion may range from 1 year or season for annually laminated (varved) sediment to more than 20 years for lakes with slow sedimentation rates. Several components of sediments provide information on factors related to lake acidification. Concentrations of polycyclic aromatic hydrocarbons (PAHs), soot particles, lead, sulfur, vanadium, sulfur isotope ratios, and mag- netic particles can be interpreted to indicate trends in atmospheric deposition of substances derived from com- bustion of fossil fuels. The pH of lake water during the past can be inferred by analyzing assemblages of diatoms and chrysophytes. Remains of chydorids (littoral crustaceans) and chironomids (midge larvae) may also provide insight into changes in aquatic biota related to acidification. Acidification of lakes and their watersheds can be inferred from changes in concentrations of common and trace metals, such as calcium (Ca), magnesium (Mg), sodium (Na), potassium (K), zinc (Zn), lead (Pb), aluminum (Al), manganese (Mn), and iron (Fe). Various disturbances of watersheds can be indicated by the common and trace metals, pollen, charcoal, and changes in sedimentation rate based on lead-210 dating. IndicatiOnS

337 of watershed disturbance should be corroborated if at all possible by thorough investigation of historical records and other studies, such as tree ring analysis. Of the above characteristics of sediment, the most information available for acid-sensitive lakes is derived from data on sediment diatom assemblages and sediment chemistry, particularly trace metals. Thus, in this chapter we emphasize these data. DIATOM AND CHRYSOPHYTE SEDIMENT ASSEMBLAGES Analyzing and interpreting sediment diatom and chrysophyte assemblages is the best paleolimnological technique available for reconstructing past lake-water pH. Investigators are using this approach increasingly to assess changes that may have been caused by atmospheric deposition of strong acids, because patterns of change in lake-water pH can be used to help determine trends in acid deposition. Diatoms and Chrysophytes as Indicators of Lake Chemistry Diatoms make up a large group of single-celled freshwater and marine algae (division Bacillariophyta). They have siliceous cell walls and are formed of two halves or valves. Chrysophytes (Chrysophyceae) are primarily freshwater plankton. In this review the term chrysophyte refers to only one family, the Mallo- monadaceae, also known as the scaled chrysophytes. Its members have flagella and an external cell covering of overlapping siliceous scales and bristles. The scales are used for paleoecological reconstructions. The distributions of diatom taxa are closely related to water chemistry (Cleve 1891, Kolbe 1932, Hustedt 1939, Jorgensen 1948, Cholnoky 1968, Patrick and Reimer 1966, 1975, Patrick 1977). For this reason, diatoms are commonly used as indicators of pH, nutrient status, salinity, and other water quality characteristics (e.g., Lowe 1974). Stratigraphic analysis of fossil diatom assemblages can be used to investigate changes in lakes resulting, for example, from shifts in climate, develop- ment of watershed soils and vegetation, local human disturbance of watersheds, and acid deposition (e.g., Battarbee 1979, 1984, Pennington 1981, Fritz and Carlson 1982, Brugam 1983, 1984, Del Prete 1972) .

338 Diatom assemblages in sediment are good indicators of past lake pH because (1) diatoms are common in nearly all freshwater habitats, (2) distributions of diatom taxa are strongly correlated with lake-water pH (Hustedt 1939, 1927-1966; Meril'ainen 1967; Battarbee lg79, 1984; Gasse and Tekaia 1983; Gasse et al. 1983; Huttunen and Merilainen 1983; Davis and Anderson 1985; Charles 1985a; Anderson et al. in press), (3) diatom remains are preserved well in sediment and can be identified to the lowest taxonomic level, (4) their remains are usually abundant in sediment (104 to 108 valves/cm3 of sediment) so that rigorous statistical analyses are possible, and (5) many taxa are usually represented in sediment assemblages (20 to 100 taxa per count of 500 valves is typical) so that inferences are based on the ecological characteristics of many taxa. Some disadvantages in using diatoms as pH indicators are that (1) diatom identification requires considerable taxonomic expertise, (2) occasionally diatoms are not well preserved because of dissolution (e.g., in some peaty and some calcareous sediments), (3) sometimes the number of taxa is low (e.g., in some bog lakes), (4) calibration data sets (the current relationship between water chemistry and surface sediment diatom assemblages) are not always available for the lake region studied, and (5) good ecological data are not always available for all dominant taxa. Other problems associated with interpreta- tion of diatom data are discussed at the end of this section. In general, the use of chrysophyte scales for pH recon- structions involves the same advantages and disadvantages as for diatoms (Smol 1985a,b; Smol et al. 1984a), except that the number of chrysophyte taxa in a sediment assemblage is in the range of one-tenth the number of taxa of diatoms and most chrysophyte taxa are euplanktonic (normally suspended in the water). The latter character- istic provides an advantage over diatoms in the study of acidic lakes because euplanktonic diatoms are usually rare or nonexistent in lakes with a pH below about 5.5 to S.8 (Battarbee 1984, Charles 1985a). In these cases, chrysophytes may be more sensitive indicators of water chemistry changes than diatoms because they live in direct contact with the open water, whereas most diatoms grow in the shallower water of the littoral zone, which may be chemically different from the open water.

339 Techniques for Determining pH Trends Several techniques based on diatom assemblages have been used to assess trends in acidification and to derive equations for inferring lake-water pH. These are de- scribed briefly below. Further descriptions and discussion of details and uncertainties associated with the reconstruction of lake-water pH are covered in depth by Gasse and Tekaia (1983), Battarbee (1984), Davis and Anderson (1985), Charles (1985a), and Smol et al. (1985) The simplest and most straightforward approach is to count sediment-core diatom and chrysophyte assemblages and prepare depth profiles of percentages of the dominant taxa. Changes in the profiles are then interpreted in light of the ecological data available on the taxa. At the present time, this is the only technique used to analyze chrysophyte scale data. Hustedt (1939) made one of the first significant steps toward establishing a more quantitative approach for using diatoms as pH indicators. He recognized the strong relationship between diatom distributions and lake-water pH and defined the following pH occurrence categories: Acidobiontic--optimum distribution at pH below 5.5 Acidophilic--widest distribution at pH less than 7 Circumneutral/indifferent--distributed equally above and below pH 7 Alkaliphilic--widest distribution at pH greater than 7 Alkalibiontic--occurs only at pH greater than 7 Assignments of diatom taxa to these categories can be based on literature references and on the distribution of taxa within waters of particular geographic regions. Changes in the percentages of diatom valves in each pH category in a sediment core can be used to estimate trends in lake-water pH. This method makes use of data on most of the taxa within a core, not just the dominant species. Nygaard (1956) took the next major step with the development of a set of indices. These indices are based on ratios of the percentages of diatom valves in Hustedt's pH categories. First, acid units and alkaline units are calculated. acid units = 5 (% acidobiontic) + (% acidophilic), alkaline units = 5 (% alkalibiontic) + (% alkaliphilic)

340 The percentages of valves in the two extreme pH cate- gories, acidobiontic and alkalibiontic, are arbitrarily weighted by a factor of 5 because diatoms in these categories are presumably stronger indicators of pH. The formulas for Nygaard's indices are acid units a = alkaline units acid units number of acid taxa alkaline units number of alkaline taxa . Of these, index a is the best predictor of current pH in acidic lakes (Merilainen 1967, Davis and Anderson 1985, Charles 1985a; Figure 9.1) and is used most frequently. Renberg and Hellberg (1982) derived a new index (Index B), also based on pH categories: (% indifferent) ~ 5 (% acidophilic) + 40 (% acidobiontic) Index B = (96 indifferent) + 3.5 (% alkaliphilic) + 108 (% alkalibiontic) Index B has advantages over index on, including the use of more information and less reliance on alkaline taxa, which are typically rare or absent in acid lakes. New indices incorporating Hustedt's (1939) pH categories have been developed by Watanabe and Yasuda (1982) and Brakke (1984). Merilainen (1967) refined quantitative techniques even further by developing an approach to predict lake-water pH from the index values. The relationship between tne logl0 of index values and measurements of lake-water pH for several lakes is determined by using regression analysis (e.g., Figure 9.1). Predictive equations are then derived directly from the slope and intercept of the regression equations. Predictive equations can also be developed from multiple linear-regression analysis of measured lake- water pH with the percentages of diatoms in each pH category (e.g., Davis and Anderson 1984, Charles 1985a, Figure 9.2). Other approaches for inferring pH involve the use of multiple regression of selected taxa and

1 0 9 8 I 7 6 5 4 341 Denmark (Nygaard, 1956) * Finland (Merilainen, 1967) O Norway ( Davis et al., unpublished ) \ * · Adirondacks, U.S.A. ([gel Prete et al., 1972; Schof ield and Ga I loway, 19 77 ) · Northern New England, U.S.A. (Norton et al., 1981 ) pH = ~.73252 (1°91 0 a) + 6.6024 *\ ~ · - ' 0~ - :~^Oo Hi- my- · O * \ ~,0^ . . 8 -amp ~ ^1: 1 ~ -4 -3 -2 -1 0 1 2 3 L°G1 0 0t 4 oo FIGURE 9.1 Logarithm of Nygaard's alpha index for surface sediment diatom assemblages versus lake surface water pH (Norton et al. 1981). Data for Norway are from R. B. Davis, F. Berge, and D. Anderson, University of Maine, Orono, unpublished data. multiple regression of principal components of taxa data sets (Davis and Anderson 1984, Gasse and Tekaia 1983). The standard error for inferred pH ranges between +0.25 and +0.5 pH units (Battarbee 1984, Davis and Anderson 1984, Charles 1985a). Usually, trends in pH curves are not analyzed statis- tically. Instead, subjective interpretations are made that account for the nature of the diatom assemblages, the error associated with the predictive techniques, evidence of sediment mixing, and other factors. This is not a problem if pH changes are great, but relatively small changes, for example, within the standard error of the predictive equation, must be interpreted cautiously, especially if the changes do not show a consistent trend Esterby and El-Shaarawi over several sediment intervals.

342 8 — J O J CD ~ 111 o Cr I 6 — 11. O in O Oh in IN to I L`J r2 = 0.94 5 — / / ~ ., 4- , ~ / ~ /e ~ / ~ ~ /. / / · , / / / / ~ / ~ · / i' ,- t-~. · ·, 4 5 6 ME aSUR E D p H 7 8 FIGURE 9.2 Predicted lake pH calculated using the equation derived from multiple linear regression of pH categories (assignments based primarily on the distribu- tion of taxa in Adirondack lakes) versus measured surface pH for 37 Adirondack lakes. The dashed lines represent the 95 percent confidence intervals for an individual prediction of pH from diatom data (Charles 1985a). (1981a,b) have developed a point-of-change technique to determine the point of maximum rate of change in a profile of dominant diatom taxa and whether the change is statistically significant. This technique has been used to evaluate taxa profiles from at least one lake (Delorme et al. 1984). The technique has also been applied to pH profiles (G. W. Oehlert, University of Minnesota, personal communication.) New statistical approaches should be developed that account for different sources of uncer tainty (e.g., Oehlert 1984). Diatom and chrysophyte data can be used to address questions such as: Has a lake become more acidic or -

343 alkaline? How great were the changes? When did they occur? What were the causes? The extent to which these questions can be answered depends on (1) the quality of sediment cores, (2) the preservation and diversity of sediment diatoms, (3) the quality of diatom slide prepara- tion and counting methods, (4) the quality of taxonomic identifications, (5) the precision and accuracy of the pH inference techniques and the applicability of the equa- tions to a study region or lakes, (6) the accuracy and precision of dating and other information on sediment characteristics, and (7) the availability of historical watershed and atmospheric deposition data. Evaluation of Lake Acidification Causes Diatom and chrysophyte data can be used not only to infer the past pH trend of a lake but in many cases to suggest the causes of the changes. There are three major potential causes of acidifica- tion of relatively undisturbed, acid-sensitive lakes in eastern North America: (1) long-term natural acidifi- cation, (2) watershed disturbances, such as logging and fires, and ensuing responses of vegetation and soils, and (3) atmospheric deposition of strong acids from distant sources. Other factors may affect lake pH, but not on a regional scale. These include nearby emission sources, discharge of factory effluent, cultural development (roads and houses, for example), land clearance for agriculture, afforestation (more common in Europe), acid mine drainage, mixing with seawater, liming, paludification, drainage of wetlands, and water-level changes such as those resulting from small man-made dams, beaver activity, or changes in climate. The last factor is probably important only when a significant proportion of a watershed is wetland and net sulfur reduction-oxidation and cation exchange processes are affected. Because these other factors have not, with few exceptions, affected the lakes evaluated in this report they are not considered further. The three primary potential causes of lake acidifica- tion are addressed below. Diatom studies of long-term acidification in both Europe and North America are also briefly reviewed. Following this is a summary of patterns of diatom changes to be expected in response to each major acidification cause.

344 Long-Term Natural Acidification Long-term trends in the postglacial development of Temperate Zone European and North American lakes have been studied using sediment diatom, pollen, and chemical analyses. These studies indicate that acidic to weakly alkaline lakes in areas having bedrock that is relatively resistant to weathering have undergone a gradual long-term acidification process. This process has been recognized at least since the studies of Lundquist (1924) and has been observed in many geographic areas. In Europe these areas include Sweden (Digerfeldt 1972, 1975, 1977, Renberg 1976, 1978, Renberg and Hellberg 1982, Salomaa and Alhonen 1983, Tolonen 1972), Finland (Alhonen 1967, Tyrni 1972, Tolonen 1967, 1980, Tolonen et al. 1985), Denmark (Nygaard 1956, Foged 196g), northwestern England (Evans 1970, Haworth 1969, Round 1957, 1961, Penning ton 1984), Scotland (Alhonen 1968, Penning ton et al. 1972), Wales (Crabtree 1969, Evans and Walker 1979, Walker 1978), Czechoslovakia (Rehakova 1983), and Greenland (Foged 1972). Fewer regions in North America have been investigated, but diatom data indicate that long-term acidification has occurred in Mirror Lake, New Hampshire (Sherman 1976); Cone Pond, New Hampshire (Ford 1984); Bethany Bog, Connecticut (Patrick 1954); Berry Pond, Massachusetts (Rochester 1978); Heart Lake, Upper Wallface Pond, and Lake Arnold in the Adirondack Mountains, New York (Reed 1982; Whitehead et al. in press; Figure 9.3); Crystal Lake, Wisconsin (Conger 1939); Lake Mary, Wisconsin (J. C. Kingston, University of Minnesota at Duluth, personal communication); Vestaberg Bog, Michigan (Colingsworth et al. 1967); Red Rock Lake, Colorado (Norton and Herrmann 1980); and on Ellesmere Island, above the arctic circle (Smol 1983; J. Smol, Queens University at Kingston, personal communication). We can make some generalizations based on these studies. The magnitude and the rate of long-term acidification vary among lakes and within lakes. Diatom- inferred pH, when it has been calculated, indicates declines from 0.5 pH unit or less to about 2.5 pH units; for example, from pH 7.5 to pH 5.0 for Upper Wallface Pond (Whitehead et al. in press; Figure 9.3). Rates of change are gradual; declines of 1 pH unit take hundreds to thousands of years. In general, lakes with the highest current pH have acidified the least, and data for lakes with pH currently above about 7.5 indicate little

345 2 m Oh 4 — u, 6 it .$ 1_ 10— 12 · ;~,~ 1 ~ 4, ~ t ¢ ., ~ ) t ', _. / ~ 'odd a. "_ ~ en_ o 0 U. WALLFACE -~ HEART LAKE - - LAKE ARNOLD . . do . -' I~,~ _7~'mp - . 1 4.5 5.5 6.5 7.5 AVERAGE INFERRED pH FIGURE 9.3 Weighted average diatom-inferred pHs for Heart Lake, Upper Wallface Pond, and Lake Arnold, Adirondack Mountains High Peaks Region. Averages were determined from index a, index B. and multiple- regression equations. Each pH value was weighted by the standard error of each predictive equation. B. P. signifies before present. Whitehead et al. in press.

346 or no long-term acidification (e.g., Linsley Pond (Patrick 1943), Pickerel Lake (Haworth 1972), Sunfish , Lake (Sreenivasa and Duthie 1973), Kirchner Marsh (Brugam 1980), South King Pond (Ford 1984)). The most rapid acidification generally occurred during the early postglacial period. Following deglaciation, the lakes were commonly alkaline (pH 7.5 or above), the major reason for the high pH being the presence of unweathered till and outwash deposits with large supplies of easily leachable base cations. Alkalinity declined after easily leached bases were removed and as soil and vegetation developed. The roles of these processes in soil and Lake development have been shown by Cracker and Major (1955), Livingstone et al. (1958), Andersen (1966), Berglund and Malmer (1971), and Jacobson and Birks (1980). _ The extent of the decline in pH and alkalinity depended largely on the characteristics of the soil, bedrock, climate, and vegetation. Acidification was greater in regions where geologic materials are more resistant to weathering, where soils and glacial deposits are thinner, where rainfall is greater, and where vegetation results in greater output of organic acids. The early period of most rapid acidification was usually followed by long periods of relative pH stability or very gradual decline (usually 1.0 pH unit or less). Some fluctuations in pH occurred and are attributed to changes in such factors as climate, watershed vegetation, and lake-level changes. For European lakes especially, other events such as seawater intrusion (connection with saltwater environ- ments) and land use changes caused by early inhabitants also affected lake-water pH. Watershed Disturbance Logging, fire, blowdown, and similar disturbances can potentially change lake pH and alkalinity, although there is evidence that these changes are often relatively minor and short lived (Rosenqvist 1978, Gorham et al. 1979, Wright 1981, Nilsson et al. 1982, Martin et al. 1984). A pattern of pH change that might be expected is an initial, rapid increase in pH as the flux of cations released from a watershed increases. This would be followed by gradual decline to below predisturbance pH as cation flux decreases because of landscape stabilization and increased uptake by rapidly regrowing vegetation (Nilsson et al. 1982, Gorham et al. 1979). Alternatively,

347 the loss of a forest canopy could change the type and amount of dry deposition, thereby altering the depo- sitional flux of acidifying compounds. Factors affecting the response of a lake to watershed disturbance include (1) the nature, extent, and severity of the disturbance; (2) past frequency of similar disturbances; (3) the influence of drainage pattern and density, geology, soil type and texture, steepness of watershed slopes, and hydrology; and (4) taxonomic composition and rate of vegetation regrowth. If initial major changes are short lived (a few years at most), they might not be detected by analysis of sediments. This could occur if the appro- priate sediment interval representing the disturbance time were not analyzed or if altered diatom and chrysophyte assemblage composition were not discernible because of sediment mixing or some other factor. Atmospheric Deposition of Strong Acids Precipitation containing excess strong acids is falling on large areas of eastern North America (Chapter 5) and western Europe. Available evidence suggests that the quantities of strong acids now being deposited are sig- nificantly greater than they were before the year 1800 (Chapter 2) and that this precipitation can cause acidification of lakes. Some authors suggest, for example, that the concentration of sulfate in many Adirondack lakes may be five times higher than before 1800 (Galloway et al. 1983, Holdren et al. 1984). Diatom-based pH reconstructions have been used to assess the effect of acid deposition on lakes in Europe and North America. For Europe, recent declines in diatom-inferred lake-water pH related to acid deposition have been reported for Sweden (Almer et al. 1974, Renberg and Hellberg 1982), Norway (Berge 1976, 1979, 1983, Davis and Berge 1980, Davis et al. 1983, Davis and Anderson 1985), Finland (Tolonen and Jaakkola 1983, Tolonen et al. 1985), Scotland (Flower and Battarbee 1983, 1985), The Netherlands (van Dam et al. 1981, van Dam and Kooyman-van Blokland 1978), and the Federal Republic of Germany (Arzet et al. 1985). Most of these studies, and some on North American lakes, were reviewed by Battarbee (1984). He concludes that Despite some ambiguous situations associated with land use changes the weight of evidence

348 favours acid precipitation as the main cause of recent acidification in the lakes so far studied. Its effect has yet to be disproved at any site and the temporal and spatial patterns of acidification within NW Europe, limited though the data are, are consistent with such an hypothesis. In all cases the onset of acidification postdates about 1800, after the development of coal as a major power source during the industrial revolution, and the later acidification of Swedish and Finnish lakes may be associated with a postwar increase in emissions from oil combustion (Otter, 1977) as well as a change in the pattern of emissions. Evaluating Causes of Recent Acidification Long-Term Acidification Declines in lake pH caused by natural, long-term acidification processes should be continual throughout the period studied. Based on studies of long-term lake histories the overall decline should be small, no more than about 0.1 pH unit in a 20-year period. There should be no major changes in taxa composition or at least no major shifts in the percentage of diatom valves assigned to each pH category. Watershed Disturbance If a pH decline were caused only by watershed disturbance and processes associated with vegetation recovery, the period of pH changes should be logically related to the time of the disturbance; the larger and more rapid the pH changes, the closer they should have been to the time of disturbance. A pattern that might be expected is a sudden change in pH followed by a gradual return to previous conditions, although there are few data to substantiate this contention. In the case of fires there may be an increase in pH followed by a decrease (e.g., Dickman and Fortescue 1984). In addition to changing the pH, watershed disturbance may affect the nutrient input to lakes. This effect, in and of itself, can cause changes in the composition of diatom assemblages. Atmospheric Deposition of Strong Acids Lake acidification can be attributed to acid deposition if (1) the lake is in a region receiving deposition of excess strong acids

349 and there is evidence of industrial pollution in the upper levels of the sediment; (2) inferred pH declines occurring after an increase in acid deposition could reasonably be expected to have taken place, based on knowledge of current water chemistry and watershed characteristics; (3) the pH decline is more rapid than could be accounted for by natural acidification processes; and (4) changes are not correlated with, and cannot be attributed to, watershed changes or other local factors. When these criteria are used to evaluate causes of acidification, several factors must be emphasized. First, more than one of the above processes may be affecting a lake simultaneously. Second, the history of each lake- watershed system and nearby emission sources should be thoroughly investigated so that all potential causes of acidification can be evaluated. Third, interpretations based on diatom and chrysophyte data should be consistent with interpretations of other sediment data, historical water chemistry and fish population data, and current models of lake acidification processes. Fourth, assess- ments of regional acidification trends should be based on studies of several lakes within a region. ASSESSMENT OF RECENT LAKE ACIDIFICATION TRENDS IN EASTERN NORTH AMERICA To evaluate recent acidification trends in eastern North America, 27 diatom data sets meeting minimum criteria were assembled and reviewed. Locations of the s lakes are shown in Figure 9.4. In addition, four lakes in Rocky Mountain National Park, Colorado, were selected as controls. Characteristics of these lakes and their diatom assemblages are summarized in Appendix Tables E.1 to E.5. The steps in analyzing the data were as follows: (1) assemble all relevant data sets, (2) evaluate pH and acidification trends by type of lake and by geographic region, (3) evaluate potential causes of the trends, and (4) draw conclusions concerning trends in acid deposition. e Data Selection The criteria for selecting lakes and diatom assemblage data were as follows:

350 am; ~=,26 ~ I ~ /57 C. 27 REGION A .20\ ~ REGION D ~ REG ION B en .~C _~ FIGURE 9.4 Locations of 27 lakes in eastern North America for which data on diatoms in sediment are available. The number next to each lake corresponds to the order in which it appears in Appendix Table E.1 and in subsequent tables. The depicted regions have been defined in Chapter 1. 1. Lake-water alkalinity less than 200 micro- equivalents per liter (peq/L). Acid neutalizing capacity is low enough so that lake pH and diatom assemblage composition could have been altered as a result of significant change in watershed characteristics or atmospheric deposition of acids. 2. Minimal or, if not minimal, well-known watershed disturbance or influence from local emission sources, industrial effluents, or other factors.

351 3. Sufficient quality and quantity of diatom data, with taxonomic identification to lowest possible level, an adequate number of time intervals analyzed, and access to primary data including diatom counts. 4. Extent to which diatom data have been related to lake-water pH and alkalinity (e.g., profiles of percent of dominant diatom taxa only versus inferred pH calculations). 5. Quality of sediment dating and the availability of data for other sediment characteristics related to acid deposition and lake acidification. Chrysophyte data were available only for Delano Lake, Jake Lake, and Found Lake in Algonquin Park (Smol 1980) and for Big Moose Lake (Smol 1985a), Deep Lake, Lake Arnold, and Little Echo Pond (J. P. Smol, Queens University at Kingston, personal communication) in the Adirondack Mountains. . Major changes in diatom stratigraphy were usually accompanied by changes in chrysophyte stratigraphy and they both indicated the same pa trend. Many of the data represented in Appendix E (Tables E.4 and E.5) are as yet unpublished. This is because inves- tigators have collected the data only recently, and their analyses are currently incomplete. Most of the data listed in Appendix E are for lakes in the Adirondack Park, New York, or in New England. In general the number of lakes is too few for making regional assessments, and the lakes are not broadly representative of each region, even of the poorly buffered lakes. All the Canadian lakes reported, except B and CS, have a pH above 6.0; a large percentage or eastern Canadian Cares, however, have pH values less than 6.0. m e nine New England lakes are among the most acidic in the region and are among the 20 lowest-pH lakes of the 94 lakes studied by Norton et al. (1981), all of which had a pH less than 7.0. With a few exceptions, the surface sediment diatom- inferred pH values agree within about 0.1 to 0.4 unit of current lake-water measurements (Appendix E, Table E.5). Trends in pH On a regional basis, only diatom and chrysophyte assemblages in the Adirondack lakes indicate significant recent acidification. The three lakes with recent measured pH above 5.5 acidified only slightly (Panther

352 Lake; Del Prete and Galloway 1983, Davis et al. in preparation) or not at all (Seventh Lake, Del Prete and Schofield 1981; Sagamore Lake, Del Prete and Galloway 1983). The pH of Big Moose Lake (Figure 9.5) and Honnedaga Lake declined about 1 pH from before 1800 to the present. The decrease in the pH of Big Moose Lake is associated with recent increases in acidobiontic taxa, including Fragilaria acidobiontica Charles and Stauroneis gracillima Hustedt, and Decreases In the indifferent, euplanktonic taxon Cyclotella stelligera Cleve and Grunow (Figure 9.6). The increase in chrysophyte taxa indicating acidic conditions, such as Mallomonas hamata Asmund and M. hindonii Nicholls (Smol et al. 1984a,b; Figure 9.7), . also indicates a lowering of pH. Diatoms and chrysophytes in Deep Lake, Lake Arnold, and Upper Wallface Pond indicate a recent drop in pH from about 5.0 to around 4.7. Shifts in taxonomic composition (e.g., Charles 1985b) suggest that increased concentrations of aluminum or other metals may have accompanied the pH decline. Trends in pH for three other lakes are more difficult to interpret. Diatoms and chrysophytes in Little Echo Pond. a bog pond, suggest a relatively small pH change. Little change in average pH over time would be expected in a highly colored acidic bog that is buffered by organic acids. However, one taxon, Asterionella ralfsii var. americana Korner (acidophilic to acidobiontic), dramati- cally increases in abundance (to >90 percent of the diatom assemblage) in the top few centimeters of the core. This may represent a natural fluctuation or may be an indication of acidification. Interpretation based on changes in the percentage of only one taxon are generally much less reliable than those based on change in the percentages of several taxa. There is no obvious pH trend for Woodhull Lake. How- ever, there is an unexplained difference of 0.7 to 0.8 pH unit between diatom-inferred pH and available measure- ments of recent lake-water pa. Del Prete and Galloway 'S (1983) analysis of a Woods Lake core indicated a gradual, though somewhat irregular, pH decrease continuing to recent times. The investigation by Davis et al. (in preparation) of a different Woods Lake core infers an increase in pH in the period 1875-1910 followed by a pa decrease from 1920 to the present. The surface sediment diatom flora suggest more acidic conditions than do the pre-1800 flora. Changes in sediment diatom assemblage composition in New England lakes indicate recent acidification (Norton

353 19821 1960. 1940 1920 us 6 ~ 1 900 A N Q 1880 1860 1840 1820 1800 ~ ·-~% ~ ~ ~ ~- 4.5 5.0 5.5 DIATOM INFERRED pH ~ O .? _. ~ > I. . - 25 . · ~ · \ · . > ~ . ~ ~ · ~ ~ 6.0 6.5 ~ 5 -10 -15 - _20 -30 -35 INDEX + SD - MR 0.28 B 033 ....... ct 0.33 ~ 95% Cl 0.40 FIGURE 9.5 Profiles of inferred pH from Big Moose Lake (core 2) based on a multiple-regression equation (MR, solid line), index B (dashed line), and index a (dotted line). Standard deviations are 0.28, 0.33, and 0.33, and the 95 percent confidence intervals are 0.40, 0.55, and 0.56 for the OR, index B. and index a equations, respectively. strength of relationship between measured and predicted pH: MR equation, pH = 8.14 - 0.041 ACB - 0.034 ACE - 0.0098 IND - 0.0034 ALK' r2 = 0.94; index B. pH = 7.04 - 0.73 loglO index B. r = 0.91; index a, pH = 6.81 - 0.70 loglO index a, r2 = 0.91. ACB, acidobiontic; ACE, acidophilic; IND, indifferent, circumneutral; ALK, alkaliphilic. Adapted from Charles (1984). Equations used for AH inference, and , ~

354 30 C: UJ ~ 20 O 10 en Al :D 20 A: ~ 10 O O 20 ~ 10 o O 60 ~ 40 uJ 20 6.0 4.5 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980 30 Diatom 1—Acidobiontic Diatom 2—Acidophilic t:: ~ nt:~t~m `1_^ir—~emno~ ~ 20 Percent acidobiontic diatoms 1 1 1 1 1 1 1 1780 1800 1820 1840 1860 1880 1 1 1 1 1 1 1 1 1 1 1 ~ 1 ~ 4.5 1980 YEARS 1900 1920 1940 1 960 FIGURE 9.6 Reconstructed pa, percent of acidobiontiC diatoms, and percent abundance of three diatom taxa, Big Moose Lake sediment core 2. Reconstructed pH is a weighted average of index a, index B. and the multiple- regression equations shown in Figure 9.2. Weighting was based on the standard error of the three predictive equations (Figure 9.5). Diatom 1, Stauroneis aracillima. . . . diatom 2, Anomoeoneis serians var. bracYsira, and diatom 3, Cyclotella stelligera, represent the dominant taxa in the core and occurred in subsets of 38 study lakes with mean pH values of 4.8, 5.9, and 6.8, respectively (Charles 1984). Percentage of euplanktonic taxa within the core followed the pattern of C. stelligera abundance very closely. Years are based on lead-210 dating (Figure 9 .5) ~ et al. 1985); however, most of the pH-trend data indicate no change or a slight decrease (e.g., in Speck Pond; Figure 9.8) in pH from pre-1800 to the present. At Ledge Pond, a total decrease of 0.6 to 0.7 unit, starting slowly during the 1800s but accelerating after 1960, has been inferred (Davis et al. 1983, Norton et al. 1985). There are fluctuations in some of the profiles that appear related to watershed events (Davis et al. 1983). The Algonquin Park lakes and Batchawana Lake have apparently undergone little change in pH in the last 200

355 of .~. _ to eGV ^0 REV 1960— Ul 1940- - lo D _ 1920 - 1~ _ 1~ _ ~ 1~ _ - 1~ _ — 1~ _ . o ~0~ _C~.~ +. 5~ ff~5. At _ ._ I I . 1 1 l ~ _ _ , . O 10 20 0 10 20 0 10 0 5 0 10 0 10 200 10 PERCENT He than 1%} FIGURE 9.7 Percentage diagram of scales of dominant chrysophytes in Big Moose Lake sediment core 2 (Smol 1985a). —0 —10 _ E Cal - - 20 ~ Ul years. This is not unexpected given that their current alkalinities are generally greater than about 50 peq/L. _ _ from a slight change by using diatom data when alkalinities are this amount or above. Lakes B and CS have apparently become more acidic in the past 30 to 50 years. Much of the cause could be local sources of sulfur emission near Wawa, Ontario, that increased output during the 1940 s and 1950s (Dickman et al. 1983, Dickman and For tescue 1984, For tescue et al. 1981, 1984, Somers and Harvey 1984). Therefore, the pH trends cannot be used to help interpret changes in long- rance transport of strong-acid precursors. However, the It is difficult to distinguish no chance in pH _ ~ . . . . . . data are useful and are included here because they provide further evidence that sediment diatom assemblages can indicate changes caused by increased deposition of strong acids. The Rocky Mountain lakes are considered to be controls. They are all low-alkalinity lakes (three with less than 60 peq/L) sensitive to increased input of strong acids, but they do not currently receive precipitation with an annual average pH lower than about 5.0 (Gibson et al. 1984). Changes in the diatom stratigraphy are relatively

356 I N F E R R ED pH 4.5 O 5.0 5j5 1978 8 ; _ - I _ LL 16 24 ~ A. ~ taxa >by ~..~) . '~ pH t '5: groups 1 1 ~ <., 1 t)--ile--~~ 1 ·/ 1/ \J 1 963 . log ~ . a) o 1945 ~ en - LL 1 906 ~ at: to 1853 1815 FIGURE 9.8 Comparison of downcore pH for Speck Pond, Maine, calculated from multiple-regression analysis of tax a and pH groups and from logl0 index a (from Battarbee 1984). Lead-210 dates have been added to published figure; data from R. B. Davis, University of Maine, Orono, personal communication. Constant Rate of Supply model. CRS refers to minor, occur gradually, and indicate no sign of acidification. In Husted Lake there is even a slight increase in the percentage of alkaliphils toward the sediment surface. The 27 eastern North American lakes for which historic pH has been inferred can be divided into two groups, based on their background (pre-1800) pH. Thirteen had a pre-1800 diatom-inferred pH of 5.3 or less, and fourteen had a pH at that time of 5.7 or greater. The two groups represent the dominant buffering systems in the studied regions: bicarbonate (HCO3) in the higher-pH lakes, organic acids and aluminum species in the lower-pH lakes. The differences in diatom assemblages are generally

357 greater between the two groups than among individual lakes within any one group, and these specific differ- ences are similar to those observed in surface sediment assemblages and discussed by Norton et al. (1981), Battarbee (1984), Davis and Anderson (1985), and Charles (1985a). One of the greatest differences in composition is that, with one exception, euplanktonic diatom taxa are present in pre-1800 assemblages from all the lakes that had a diatom-inferred pH above 5.9 and they are present only rarely in assemblages of lakes that had a pH less than 5.4. The exception is Asterionella ralfsii var. Americana, which may be one of the few euplanktonic taxa in lakes with pH < 5.5 in eastern North America (e.g., Sweets 1983). Recent studies of surface sediment diatom assemblages of lakes in the Killarney region of Ontario show that euplank sonic Cyclotella species are common in lakes with a pH less than 5.5. The role of morphometric factors may be an important reason for the difference in occurrence of euplanktonic diatoms in these lakes compared with others in the northeastern United States (M. Taylor and H. Duthie, University of Waterloo, personal communication). The important conclusions from the distribution of pre-1800 inferred pH values are that (1) the lakes considered in this study were about equally divided overall (but not by geographic region) on the basis of the pre-1800 buffering system and (2) the number of lakes with a pre-1800 diatom-inferred pH value less than 5.5 suggests that these types of lake were relatively common in the Adirondack Mountains and New England before the Industrial Revolution. The diatom-inferred pH profiles examined for this study can be classified into four patterns, although several showed fluctuations that were probably caused by natural variation in diatom composition, watershed events, or inconsistencies introduced by pH reconstruction method- ologies. The patterns are the following: (1) background pH greater than about 6.0 and no overall trend in pH, (2) Harvey Papua acne an pa Prom the range of 5.7-6.0 to less than 5.0 (some pH declines start in pH ranges above 6.0), (3) pH about 4.8 to 5.2 throughout entire profile, and (4) decline from the range of pH 4.8-5.3 to as low as about 4.5. The biggest changes in pH and sediment core diatom composition occur when diatom-inferred pH decreases from a pH above approximately 5.5 to a pH below 5.5 (pattern 2). A smaller pH drop was observed for lakes with a _ _, _ ~ _ . ~ . ~ _ ~ · ~

358 pre-1800 pH of less than 5.5 (pattern 4). The same pattern of decline in diatom-inferred pH for currently clearwater lakes from about pH 5.0 to pH 4.5 or above has been observed in Norway (Holmvatn, Davis et al. 1983; Hovvatn, Davis and Berge 1980; Nedre Maalmesvatn, Davis and Anderson 1985), Finland (Kairajarvi and Val~eislampi, Tolonen et al. 1985), and Scotland (Loch Enoch, Battarbee 1984). Inferred pH declines in lakes with a low back- ground pH must be interpreted with caution. Even a slight decline in pH could be associated with a significant decrease in acid-neutralizing capacity and vice versa (Chapter 7). This is because the change in alkalinity per pH unit is greater in the pH range 4.5 to 5.0 than the range 5.5 to 6.0 because of the logarithmic nature of the pH scale. Changes in Water Chemistry Associated with pH Trends Three important questions arise with respect to the above discussion: 1. What was the chemistry of lakes that apparently had a pre-1800 pH less than 5.5? 2. In particular, what anions were primarily associated with the low pH? 3. What specific chemical changes could have occurred in these lakes in response to increased deposition of strong acids, and how would those affect nydrogen-ion concentration and alkalinity? In regions with low-alkalinity waters that receive precipitation that is not acidic or is only slightly so, (e.g., northeastern Minnes;c)tn once wi Knin Urn Ontario, and Labrador; Wright 1983), clearwater lakes with pH less than 5.5 are rare. However, in these regions and in regions now receiving excess acid deposition there are lakes that have pH less than 5.5 and are moderately to highly colored (platinum-cobalt units > 25) (Lillie and Mason 1983, Gorham et al. in press). These lakes are typically associated with peatland vegetation and are only slightly influenced by groundwater input. Plants such as Sphagnum can remove cations from surrounding water (Clymo 1963), and decay products of plants yield organic acids, which may or may not be colored (e.g., Vitt and Bayley 1984). These organic acids can account

359 for a significant portion of total anions in lake water and cause pH values to be less than 5.0 (Oliver et al. 1983). It seems probable, therefore, that organic acids were an important cause of acidic conditions in many lakes in acid-sensitive areas before 1800 (Patrick et al 1981). Also, if pH was about 5, aluminum concentrations may have been moderate at that time, although much of the aluminum would have been complexed with organic compounds (Driscoll et al. 1984). . As low-pH (5.0 to 5.5) lakes became more acidic, changes in aluminum and organic acid concentrations were probably the most important ones affecting pH and alkalinity. The output of dissolved aluminum from these lakes' watersheds should have increased with increases in inputs of strong acids (Almer et al. 1978, Johnson et al 1981, Driscoll et al. 1984), except perhaps for seepage lakes in outwash deposits. This is consistent with the fact that most of the lakes examined for this report that have a current pH less than 5.5 have concentrations of total aluminum greater than 100 to 200 ng/L; the value in some lakes is as high as 600 to 700 ng/L (Appendix Table E.3). The increase in percentages of diatom taxa such as Stauroneis gracillima and Fragilaria acidobiontica toward the top of some Adirondack lake sediment cores suggests that recent increases in aluminum concentration may have occurred. In the Adirondacks these taxa are abundant only in lakes with pH less than 5.5 and total aluminum concentrations greater than 200 peq/L (Charles 1985a,b). Also, computer-modeling efforts based on knowledge of current lake-water chemistry and diatom-inferred pH trends suggest that the concentration of aluminum in Big Moose Lake increased as pH decreased (Charles 1984). Increased aluminum loading could lead to reduced acid- neutralizing capacity and reduced organic acids and water color. A relatively minor decline of pH in the range 4.5 to 5.5 can cause a major change in aluminum speciation with a substantial increase in total aluminum charge (Driscoll et al. 1984), resulting in reduced acid- neutralizing capacity. In this way, aluminum acts as a buffer (Johannessen 1980), minimizing pH change as a result of increased acid loading. Increased concentration of anions is balanced more by increased positive aluminum charge than by hydrogen ion. Decreasing pH and increasing aluminum concentration could cause a decrease in organic acids and water color. This is supported by theoretical .

360 analysis, laboratory studies, and observations that water color has decreased over time in lakes that have apparently acidified recently (Schofield 1976, Almer et al. 1978, Dickson 1980). Also, studies of limed lakes indicate that as alkalinity increases, color increases and concentration of aluminum decreases (Hultberg and Andersson 1982, Yan 1983, Wright 1984). Analysis of the data discussed here suggests that when significant lake acidification has occurred in acid sensitive regions of eastern North America, it has followed two major patterns: (1) a decrease in pH of currently low-color lakes from about 5.7-6.4 to 5.0 or less and changes in alkalinity owing initially to loss of bicarbonate alkalinity and later to an increase in concentration of positively charged aluminum species; and (2) a decrease in pH from approximately 4.8-5.2 to 4.5 or above, with almost all the associated reduction in acid- neutralizing capacity caused by increases in aluminum and decreases in organic acid concentrations. In reality, patterns of pa decrease probably represent a continuum, but the above cases are likely typical of most responses. Evaluation of Trends in Acid Deposition Trends in diatom-inferred pH can be used to evaluate hypotheses concerning trends in acid deposition for eastern North America. For example, one can test the null hypothesis that atmospheric deposition of strong acids has not increased sufficiently in the past 100 years to cause significant pH and alkalinity changes in lakes with low buffering capacity. If the hypothesis is true, no evidence should exist for acidification of any type of lake in any region that cannot be accounted for by natural acidification, response to watershed changes, or another local factor. In order to reject this null hypothesis: 1. There should be evidence of decreasing pH (acidification) in poorly buffered lakes that cannot be accounted for by natural causes or response to watershed changes. 2. The above decreases should have been relatively rapid and have occurred sometime after the earliest date that independent evidence suggests acid deposition could have increased.

361 3. Decreases in lake-water pH for lakes with similar pre-1800 diatom-inferred pa should have been proportional to acid loading. For most regions in eastern North America acid deposition should have increased the most in areas that currently receive the largest loading. The quantity and the quality of diatom and chrysophyte data now available are insufficient to permit definite conclusions to be drawn regarding all aspects of the hypothesis for any geographic area. However, there is strong evidence for certain aspects. First, diatom profiles for some Adirondack and New England lakes indicate recent acidification, with the greatest pH decreases in the Adirondack lakes. Big Moose Lake represents the best case of an acidified lake in which increased acid deposition is the only reasonable cause to which recent acidification can be attributed. Diatom related changes in other lakes, such as Upper Wallface Pond and Deep Lake, are greater than would generally be expected from natural variation and are also probably due primarily to increased acid deposition, although watershed disturbance may play a minor role. There are additional data for Adirondack lakes indicating recent acidification, but either the cores were not dated or the role of watershed disturbances is unclear. These data obviously cannot be used to reject the null hypothesis, but they are consistent with the alternative hypothesis. For lakes for which adequate data are available (see Appendix E), decreases in pa were relatively rapid and occurred since about 1900, when acid deposition in the Northeast probably began increasing (Chapter 2). Changes in the Adirondacks appear to have been greater than in New England. The Adirondacks receive the greater amount of acid deposition. Thus, testing of the above hypothesis using available diatom data for the Adirondacks and New England indicates that deposition of strong acids has increased sufficiently in the past 100 years to cause a decline in lake-water pa. Data are insufficient to permit evaluation of trends in other regions, such as eastern Canada, now receiving acid precipitation. The maximum calculated pH decline for all lakes evaluated was about 1 pH unit (pH range 5 to 6) since 1900, which is similar to observations for acidified European lakes (Battarbee 1984), and the onset of rapid acidification began after 1940. However, studies of the

362 sediments of more lakes will be necessary before the magnitude and timing of recent acidification in regions can be estimated with any certainty. This research is now under way. Ongoing Studies Several lake acidification studies are now under way that include analysis of diatom and chrysophyte assem- blages as a means of reconstructing lake-water pH. The largest is the Paleoecological Investigation of Recent Lake Acidification (PIRLA) project funded by the Electric Power Research Institute. Five to fifteen lakes in each of four geographic regions (Adirondack Mountains (New York), New England, northern Minnesota-Wisconsin- Michigan, and northern Florida) have been cored and the sediments are being analyzed for diatoms, chrysophytes, and other characteristics. Patterns of pH change in lakes in these regions will be related to patterns of acid deposition to determine extent and important causes of lake acidification. A similar type of study of Rocky Mountain lakes has been completed (Baron et al. 1984, D. R. Beeson, U.S. Park Service, Fort Collins, Colorado, personal communication). Ronald Davis and Dennis Anderson of the University of Maine and other investigators are completing analysis of diatom assemblages from sediment cores of three Adirondack lakes (Panther, Sagamore, and Woods) and nine New England lakes (the first nine New England lakes listed in Appendix Table E.1). Studies of surface sediments and cores are also being conducted on lakes of the Pocono Plateau in Pennsylvania (John Sherman, Academy of Natural Sciences of Philadelphia, personal communication). The relation- ships between surface sediment diatom assemblages and limnological characteristics in eastern Canada are being studied at the University of Waterloo (at least 40 lakes, Hamish Duthie and Mark Taylor, personal communication) and at the Freshwater Institute in Winnipeg, Manitoba (about 25 lakes; Geoff Davidson, personal communication). Equations for predicting water chemistry characteristics will be developed and used in future pH reconstruction studies. Cores from lakes in the Matamek region of Quebec are being investigated (H. Duthie, University of Waterloo, personal communication), and investigations of pH histories of lakes in the Wawa region of Ontario are .

363 continuing (Michael Dickman, Brock University, personal communication). The Long Range Transport of Atmospheric Pollutants program, coordinated by F. C. Elder at the Canada Centre for Inland Waters, Burlington, Ontario, is continuing in Nova Scotia and in the Algoma calibrated watershed. A study of diatom-pH-alkalinity relationships and sediment stratigraphy of Sierra Nevada lakes began in the fall of 1985 (R. Holmes, University of California, Santa Barbara, personal communication). Chrysophyte assemblages in surface sediments of many lakes in eastern Canada are being studied also (J. P. Smol, Queens University at Kingston, personal communication). This will increase the ecological data base for chrysophytes so they can be better used in stratigraphic studies to interpret past lake changes. UNCERTAINTIES IN THE INTERPRETATION OF SEDIMENT DIATOM AND CHRYSOPHYTE ASSEMBLAGE DATA AND RECOMMENDATIONS FOR FURTHER RESEARCH Diatom and Chrysophyte Taxonomy One of the most important sources of error in diatom and chrysophyte analyses is incorrect identification of taxa. The need to resolve taxonomic problems and for better communication and more consistency in taxonomic designations among investigators is great. Comparing data among investigators is often difficult because different synonyms are used and there is inconsistency in how taxonomic levels are split or lumped. One of the major potential uses of diatom and chrysophyte analyses is to compare data among many types of lake/watershed systems in many parts of the world. More consistency in taxonomy and data presentation would aid greatly in these comparisons. Recommendation: Further taxonomic studies should be made of difficult groups and unidentified taxa, and more floras of diatoms from acidic and circumneutral lakes should be published. In addition, communication among investigators working with acid-water floras should be improved with the goal of better agreement on taxonomic categories and data presentation.

364 Diatom and Chrysophyte Physiology There are some major questions concerning distribution of diatoms and chrysophytes in acidic lakes. Answers to these questions would aid in better understanding the relationships between changes in assemblages and acidification of lakes. Recommendation: Culture and bioassay studies (e.g., Gensemer and Kilham 1984) should be performed to address at least the following questions: 1. Why do euplanktonic diatoms appear, in many regions, to be generally limited to lakes with a pH 5.8? 2. What are the physiological mechanisms underlying the strong relationship between diatom and chrysophyte distributions and factors related to lake-water pH? 3. What are the roles of aluminum, silicon, and organic materials in determining the occurrence of taxa? 4. What interactions occur between factors related to pH and those, such as phosphorus and nitrogen concentrations, related to trophic state? Diatom and Chrysophyte Ecology Interpretation of diatom data is limited by a lack of complete knowledge of diatom autecology, including infor- mation on habitats, seasonality, and relationships to water chemistry characteristics. More information on how diatoms and chrysophytes interact among themselves and with other organisms would help to explain seasonal and long-term successional patterns. Such information would also contribute to our understanding of how important factors other than those related to pH are in causing variation in assemblage composition. Recommendations: 1. Conduct more field and laboratory studies on the spatial and temporal distributions of diatoms and chrysophytes in different habitats within and among a variety of geographical regions. In particular these should be designed to establish (a) the distribution of individual taxa along specific environmental gradients and (b) the relationship between euplanktonic and peri- phytic diatoms and how their relative abundance is affected by changing conditions within a lake ecosystem, such as changes in alkalinity, trophic state, color, dissolved organic carbon, transparency, aluminum, trace metals, water level, substrate, and other biota.

365 2. Study diatom and chrysophyte changes in lakes that are experimentally acidified or limed. Sediment Processes Transport of diatoms and chrysophytes to lakes (e.g., Battarbee and Flower 1984) and within lakes, such as through differential transport of taxa (Sweets 1983) and sediment focusing, and changing patterns of processes within sediments, such as dissolution and mixing, affect The degree to which sediment assemblages represent those living in a lake. The extent to which these factors influence historical reconstructions is difficult to determine. Recommendations: 1. Perform more field work to determine how diatoms and chrysophytes are transported to lake sediments and which taxa or types of diatoms tend to be overestimated or underestimated within the sediments. Statistical analyses may be needed to evaluate the sensitivity of reconstruction techniques to this factor. 2. Perform more field and laboratory work to learn the mechanisms of diatom and chrysophyte dissolution and to evaluate its influence on reconstructions. Although dissolution usually has a negligible effect on assemblages (mostly only those in peaty or calcareous sediments) it can be important, especially when it influences the percentage of dominant taxa. 3. Investigate further the influence of sediment mixing on reconstructions. This can be and has been done by studying mixing processes, such as physical transport and bioturbation, within lakes and in the laboratory; use of mathematical models incorporating dating and other data; comparison of stratigraphy of varved versus non- varved sediments; comparison of long-term plankton and sediment-trap assemblages with sediment core assemblages; and the addition of large quantities of foreign marker particles to a lake followed by subsequent study to learn how they are mixed within sediments. Inference Techniques The precision and accuracy of current inference techniques, while generally good for most lakes, could be improved to give better predictions for a wider variety of types of assemblages. Also, quantitative inference

366 techniques are currently limited to determination of pH. Changes in other water chemistry parameters, such as alkalinity, color, aluminum concentration, and habitat characteristics, currently cannot be quantitatively reconstructed, although some qualitative interpretations can be made. Recommendations: 1. The size and quality of data sets of diatoms, chrysophytes, and lake characteristics should be improved. More lakes in more regions should be studied, and as many lake characteristics as are relevant to diatom and chrysophyte ecology should be measured frequently and over long periods of time. This work could be coordinated with other limnological studies. All data collected should be as compatible as possible with other calibrational data sets. 2. Inference techniques should be improved by trying new approaches. The most promising is to develop transfer functions that are multiple-regression equations using percents of selected taxa. The taxa chosen should cor- relate better statistically than those now used with lake characteristics to be inferred. To be widely applicable these equations need to include many taxa (more than 50), which is not possible today because of the limited size of calibration data sets. 3. Inference techniques should be developed for lake characteristics other than pH, such as alkalinity, color, aluminum content, and total organic carbon (e.g., Davis et al. 1985). Statistical Analysis It is often difficult to determine if diatom-inferred pH changes within a core are statistically significant because the numbers have not been subjected to statistical analysis. Recommendation: Full use should be made of existing statistical procedures, and new techniques should be developed to evaluate and express the following: 1. The precision error (standard error, standard deviation, 95 percent confidence interval) associated with individual inferred pH values. To date, error analysis has been based solely on the regression analysis used to calculate predictive equations from the calibra- tion data set. They do not account for characteristics of assemblages in sediment cores being analyzed. Error

367 estimate procedures should be developed that account for (a) variation in the calibration data set, including the pH measurements, and (b) characteristics of core assem~ blages, including diversity, the percentage of valves and diatom taxa that are used in the pH inference calcula- tions, and the probability that taxa have been placed in the correct pH category. 2. The existence and significance of trends and points of change within individual pH profiles. This type of analysis can be accomplished by using a variety of existing techniques, including regression and change- point analysis (e.g., Esterby and El-Shaarawi 1981a,b). 3. Variability among the pH values inferred by using a variety of indices. It is often best to present several inferred pH curves derived by using a variety of tech- niques, but this can sometimes be confusing. The alterna- tive is to use only one or two that are considered to be the best, but this involves subjective judgments and the temptation to use the curve that best fits an inves- tigator's interpretation. Objective criteria for selecting and presenting data are necessary. 4. The similarity of patterns expressed by more than one diatom-inferred pH profile from either the same or several cores. The data can be analyzed statistically by using regression analysis or Monte Carlo simulation. Acidification Processes and Selection of Study Sites In some cases, diatom and chrysophyte data sets may not yield much information about lake acidification or atmospheric deposition because (1) the study lakes are well buffered and have changed little in alkalinity or diatom composition even though atmospheric loading of acids may have increased substantially; (2) there was a major watershed disturbance, such as fire, logging, and cultural development, that obscures the relationship between acid deposition and the watershed changes as causes of acidification; and (3) the lake sediments are very mixed (physical mixing or bioturbation) or there is some problem with the core or dating of the sediment. Recommendations: 1. Choose study lakes that have a l current alkalinity between about -10 and 50 peq/L. These lakes would probably have had low enough buffering that their alkalinity could have changed enough under the influence of acidic precipitation to cause a change in

368 diatom assemblages. Changes in pH in lakes with lesser or greater alkalinities would be less likely to be observable because of greater buffering. 2. If possible, choose lake/watershed systems that have not been disturbed significantly within the time period of interest. If this is not possible, choose systems with the least amount of disturbance and for which a good historical record of the changes exists. 3. Investigate several sites within a region. Choose sensitive systems with different watershed, morphologic, and hydrologic characteristics. 4. Investigate many sediment characteristics within cores. The additional data aid greatly in evaluating the nature of lake changes and assessing causes and mechanisms of acidity change. 5. Select study systems and evaluate results in the context of current lake-acidification models. Interpreta- tions of diatom and chrysophyte data with respect to changes in acid deposition (or other factors influencing lake-water acidity) should account for limnological processes (e.g., sediment buffering, sulfate reduction in the hypolimnion and sediments, and change in organic acids) and watershed biogeochemistry changes (e.g., aluminum leaching, sulfate adsorption on soils, and cation leaching). Diatom inference data can be used to evaluate acidification hypotheses and models (e.g., the amount by which pH or alkalinity should have changed over time given certain model assumptions). After minimal lake selection criteria are met, base the final selection of study lakes on the hypothesis to be tested. Ideally, paleoecological, limnological, and watershed studies should be integrated. ACKNOWLEDGMENTS Donald Charles thanks Dennis Anderson, Klaus Arzet, David Beeson, Ronald Davis, L. Denis Delorme, Anthony Del Prete, Hamish Duthie, Jesse Ford, John Kingston, John Smol, and Kimmo Tolonen for providing access to their unpublished data. They also provided interpretation of their results and many useful comments on the manuscript. Jill Baron, Charles Driscoll, Jesse Ford and Stephen Norton contributed water chemistry and other information on many of lakes mentioned in this section. Some of the unpublished Adirondack region data were obtained as part of research efforts supported by the Electric Power Research Institute, Palo Alto, California.

369 THE CHEMICAL STRATIGRAPHY OF I,AKE SEDIMENTS The unpolluted atmosphere consists largely of the gases nitrogen (78 percent), oxygen (20 percent), argon (1 percent), water, and carbon dioxide with variable but trace amounts of other gases, such as ammonia, carbon monoxide, mercury, methane, and nitrous oxide. In addition, natural processes inject particulate organic and inorganic material into the atmosphere, matter that ultimately sediments out. Scientists have long known (e.g., Smith 1872) that human activities have resulted in alteration of the chemistry of the atmosphere with respect to both gaseous and particulate material. Nonurban regional pollution of the atmosphere was clearly iden- tified as occurring as early as the 1900s in Scandinavia (e.g., Oden 1968), in the 1950s in Great Britain (Gorham 1958) and the United States (e.g., Junge and Werby 1958), the 1960s in Europe (Oden 1968), and in the 1970s in Canada (Beamish and Harvey 1972). Gaseous pollutants (sulfur dioxide, sulfur trioxide, nitrous oxide, and nitrogen dioxide) were the earliest recognized pol- lutants. Significant pollution by airborne metals of regional atmospheres in the United States was suggested by the pioneering work of Lazrus et al. (1970), who measured lead, zinc, copper, iron, manganese, and nickel in precipitation over a period of six months. Whereas the collection and analytical techniques of this study may be challenged in terms of implied concentration and deposition rate of pollutants, the geographic trends are clear. The northeastern United States was receiving significantly more trace metals through precipitation than was the Southeast, Midwest, or West. Lakes receive a flux of gaseous, liquid, and particu- late material directly from the atmosphere as both dry and wet deposition. Additional terrestrial fluxes of these materials derived from the atmosphere, altered in quantity and quality by terrestrial processes, may also reach the lake. Physical, biological, and chemical processes in the lake sequester a fraction of the flux of these pollutants in lake sediments. In the absence of long-term, regional studies of atmospheric fluxes and deposition of pollutants the chemistry of lake sediments may serve as one of the few surrogate indicators of atmospheric pollution. However, the faithfulness of reproduction of the atmospheric deposition signal by sediments is highly variable, depending on the constituent being deposited and the

370 ecosystem characteristics. Important variables include water chemistry, focusing of sediment, resuspension of sediment, the flux of material from the drainage basin, and organic productivity within the lake. After initial deposition, sediment may be resuspended and resedimented (not necessarily in the same place) or partially mixed vertically with older (or younger) sediment by inorganic or organic processes; it may be chemically altered, too, by changing overlying water conditions (e.g., acidifica- tion and eutrophication), deoxygenation within the sedi- ments, sulfate reduction within the sediment (producing hydrogen sulfide and precipitation of otherwise mobile metals), and so on. Thus, the chemistry of a particular vertical increment of sediment is determined by the following: 1. The chemistry of the material deposited from the atmosphere as well as the chemistry of the material eroded from the drainage basin, some of which may be resuspended from other parts of the lake and deposited at the site. Redeposited sediment places older material with possibly different characteristics in a younger interval. 2. Net sedimentation of material derived from within the lake, including organisms (diatom valves, inverte- brate remains) and resuspended sediment. 3. Additions to, losses from, or relocation of inorganic and organic matter within the sediment after deposition. These changes are generally greatest at or near the sediment/water interface for oligotrophic lakes and decrease abruptly to insignificance at a depth of between S and 10 cm. (See Carignan and Tessier 1985, Holdren et al. 1984.) Thus, some of the sediment within an interval will not have the same age as the interval. The presence of laminated sediments in some lakes indicates that at least in situ mechanical redistribution of deep water sediment after one year is essentially nonexistent for a few lakes. A lake-drainage basin in a long-term steady state typically should have sediment chemistry (except for water content) that is relatively constant below a depth of no more than about 10 cm. However, changes in focusing of sediment (Dearing 1983) may result in long-term changes in sediment chemistry (Hakanson 1977). Also, lakes with greatly increased deposition rates caused by accelerated erosion and deposition with no accompanying

371 change in sediment chemistry have been identified. Thus, invariant chemistry is not a foolproof indication of steady state (Hanson et al. 1982). Changes in the above factors affecting a single chemical component result in reciprocal percentage composition changes for all other constituents in the sediment. The concentration of one constituent is generally not an independent variable. However, deposition rates for the other constituents would be unchanged by an independent change in one com- ponent, unless there were a chemical link between the two. Mechanical disturbance of the Earth's surface by activities such as deforestation, tillage, and construc- tion results in accelerated erosion of detritus from drainage basins and increased lacustrine sedimentation rates. The chemistry of the deposited material may (Davis and Norton 1978) or may not (Hansen et al. 1982) change. Coincident with or later in time than the disturbances is the increased injection of particles into the atmosphere, followed by increased atmospheric deposition. Although the composition of the particulate matter may differ from place to place and time to time, the material may be indistinguishable from normal sediment. If one can establish chronology within the sediment stratigraphy, one can evaluate the deposition rate (absolute net flux) of material at a site at the lake bottom. The distinction between deposition rate (i.e., grams per square centimeter per year) and sedimentation rate (i.e., centimeters of sediment per year) needs to be emphasized. The latter is controlled largely by water content and organic content, which are, in turn, a func- tion of a number of variables, especially lake pH. Lakes with lower pH (oligotrophic to mesotrophic) tend to have profundal sediments with higher organic content (Norton et al. 1981) and may have high sedimentation rates but low deposition rates. Modern sediment deposition rates of metals from atmospheric deposition can be evaluated approximately if one knows (1) the average background deposition rates for a particular metal; (2) the increases in the gross (bulk) deposition rate, caused by increased terrestrial erosion and other factors of all elements combined, and (3) the modern deposition rate for the particular constituent of interest. Only five to eight years ago did researchers begin to assess deposition rates accurately. This was made possible because of advances in the accurate dating of lake sediments.

372 Any particular interval of nonlaminated sediment, when collected and analyzed, is a mixture of older, contem- poraneous, and younger material. Laminated sediment may contain older and contemporaneous, but not younger, par- ticulate sediment. The age of the interval is generally thought of as the age of deposition, not necessarily as the age of the material. (Chemical diagenetic processes deny even this assumption.) The absolute age of sediment intervals can be established using varves (e.g., Tolonen and Jaakkola 1983), radionuclides associated with atmo- spheric nuclear weapons testing (especially cesium-137) (Pennington et al. 1973), and the naturally occurring radionuclide lead-210 (e.g., Appleby and Oldfield 1978). Dating with cesium-137 is imprecise in lakes suitable for studying the effects of acid rain (Longmore et al. 1981, Davis et al. 1984). It is useful in dating sediments in lakes with higher pH, higher sedimentation rates, and more abundant clay minerals, which fix the cesium-137 in the sediment. Lead-210 is the method of choice, but the method is not without uncertainty. Lead-210 chronology is, however, commonly in agreement with otner dating techniques of known validity, such as carves. Errors in collection (especially in coring techniques, such as gravity coring (Baxter et al. 1981) and small- diameter push coring (Hongve and Erlandsen 1979)), sub- sampling of the sediment, dating of the sediment, chemical analysis of the sediment, and drainage-basin and lake- basin processing of atmospheric pollutants make it diffi- cult to reconstruct precisely historic atmospheric deposition rates from net deposition rates. However, in lakes with little or no disturbances other than changing atmospheric chemistry, trends can be established. In addition, if mixing of sediments is minimal the chronology of trends can be established. Furthermore, proportional changes in the atmospheric deposition rate of certain immobile constituents, such as lead and vanadium, can be estimated. Smelting, burning fossil fuels, and other human activities inject a variety of gaseous, liquid, and particulate materials into the atmosphere that differ in composition and physical characteristics from soil dust. These materials include trace metals, such as lead, zinc, vanadium, selenium, and arsenic (Taylor et al. 1982, Galloway et al. 1982), fly ash (Griffin and Goldberg 1981), soot (Renberg and Wik 1984), magnetic spherules (Oldfield et al. 1981), organic combustion products such as polycyclic aromatic hydrocarbons (or more simply,

373 PAHs) (Heit et al. 1981), and excess sulfate, nitrate, and metals with different isotopic ratios from those that occur in soils (Nriagu and Harvey 1978). If the rates of atmospheric deposition of these substances are sig- nificantly different from natural (background) deposition rates, trends in deposition in sediments may be observed. The types of data gathered for sediment chemistry are variable but include the following: 1. Concentration of a particular constituent per unit mass of wet (least common), dry, or ignited (most common) sediment. The last-named involves total digestion of the sediment (see Buckley and Cranston 1971). 2. Concentration of a particular constituent in a particular fraction of the total sediment, such as the strong-acid-soluble fraction and the ammonium acetate- exchangeable fraction. These derived values are defined operationally, and methodological problems (see Campbell and Tessier 1984) rule out rigorously interpreting data within a sediment core and comparing sediment cores from different lakes (see Reuther et al. 1981). The most common way to represent chemical data is as concentration per unit volume or mass of sediment. This parameter, however, is not an independent variable; for any particular component it is inversely related to the concentrations of all other constituents. Even trends in concentration of some anthropogenically linked constitu- ent, such as lead derived from the atmosphere, could be diluted or enhanced in the sedimentary record by increases or decreases in the bulk sedimentation rate. In the absence of accurate age dating of sediment, concentration trends are the optimal representation of chemical data from cores. The data are commonly shown as departures from average background values. Alternatively, various types of enrichment factors (see Galloway and Likens 1979) may be used. A constituent of interest is normalized to back- ground concentrations, or a constituent is normalized to an element whose concentrates are relatively unaffected by anthropogenic activity. This approach is valid if the normalizing parameter is constant. If sediments can be dated we can calculate an absolute net flux for a constituent at a given locality at the lake bottom. For a lake receiving a constant supply of detritus it is possible to determine deposition rates for atmospherically derived substances as a bulk rate that

374 either includes autochthonous and allochthonous sources or isolates only the atmospheric component. It is clear from many studies in which multiple cores from the same lake have been studied that only fortuitously would a deposition rate at the lake bottom precisely mimic the atmospheric deposition rate. Dillon and Evans (1982) and Evans et al. (1983) have developed the concept of a whole-lake burden for con- stituents that are atmospherically derived, not delivered to the lake from the terrestrial part of the drainage basin, and not exported from the lake. Using numerous widely scattered cores, they demonstrated that the areal burden for lead did not vary appreciably from lake to lake in a small region. The logical extension of this approach is to find the representative sediment record that is characteristic of the total anthropogenic burden for the region and use it to reconstruct directly the history of atmospheric deposition. This can be done only for chemically immobile elements, such as lead, and in sedimentary environments with little mixing. Secondary effects of acid deposition may affect the sedimentary record. Effects include adsorption of chemicals by sediment, such as sulfate adsorption analogous to sulfate adsorption in soils (Johnson and Cole 1977), Resorption or increased mineral weathering caused by lowered pH (Schindler et al. 1980, Hongve 1978, Oliver and Kelso 1983, Kahl and Norton 1983, Norton 1983), reduced organic decay (Andersson et al. 1978), and altered diagenesis, for example, increased sulfate reduction followed by sulfide precipitation within either the sediment or the water column (Schindler et al. 1980). Initial deposition of constituents clearly can be modified by postsedimentation processes. As previously discussed, these effects probably become ineffective below a sediment depth of 5 cm or less. THE SEDIMENT RECORD As a result of human activity numerous elements are injected into the atmosphere. If the return flux of these elements approaches that of natural processes, detection of the perturbation becomes possible. For abundant elements, such as iron, manganese, calcium, and aluminum, natural sediment fluxes in lakes are probably in excess of atmospheric fluxes. However, processes related to various human activities cause atmospheric

375 fluxes of certain acid precursors and trace metals that are in excess of natural fluxes. m e relationship between local point sources of pollution and enrichment of these various substances in sediments of adjacent lakes is well established. The most classic example of a point source is the smelter at Sudbury, Ontario, that emits several pollutants associated with the processing of nickel and copper ore. Nriagu (1983), Nriagu and Won g (1983), and Nriagu et al. (1982) found that in dated cores from lakes proximal to the smelter the variations in concentrations (and also deposition rates) for arsenic' selenium, nickel, copper, zinc, and lead paralleled the history of ore production at Sudbury. Figure 9.9 demonstrates the sharp rise in the concentrations of arsenic in the recent sediments in a number of these lakes. The source of pollutants could be identified unambiguously as the Sudbury smelter. Similarly, Crecelius (1975) demonstrated a correspondence between arsenic in sediment and the chronology of opera- tion of a copper smelter that emitted arsenic-rich dust 35 km upwind from the lake. Other studies of lake sediment chemistry have indicated that lakes may receive excess metal inputs from multiple pathways. For example, Tolonen and Merilainen (1983) suggest that both atmospheric inputs and direct inputs from municipal sewage have contributed to increases in the accumulation rates of chromium, copper, and zinc in some Finnish lakes. Skei and Paus (1979) cite evidence suggesting that a saltwater fjord (Ranafiord) receives excess lead, zinc, and copper from atmospheric inputs and direct surface inputs, both related to nearby mining activities. Early studies performed before the importance of long-range transport of atmospheric pollutants was fully recognized presented interpretations of sediment records based on atmospheric and direct inputs from local sources For example, Edgington and Robbins (1976), using lead-210- dated cores from southern Lake Michigan, attempted to model the sediment record (for lead) in light of local atmospheric emissions as well as inputs from rivers. They concluded that fluxes of lead from anthropogenic sources overwhelmed fluxes from natural sources by nearly an order of magnitude. Downstream, in Lake Erie, Nriagu et al. (1979) concluded that the major portion of the lead, zinc, copper, and cadmium in the modern sediments was derived from direct surface inputs to the lake, not from the atmosphere. These reports and many others like .

376 Reward Lam 2o1 ~ 30 ~ ' E 1 - ARSENIC CONCENTRATION(~9/9) O S 10 15 20 0 5 10 AS 20 °! ' ~ °1 ' ~ ' } ao- l 20 - 30- 0 20 40 60 ~ I 1 1 1 f f mllllon W— 80 A 0 20 40 t;O 80 (Io 101 ~ , ~cFer~~ ~— 20 30 ' 40 ~ 50~ 60J Am r a 60' Lchn Law ARSENIC CONCENTRATION (~9~) ot 20 40 t5O 80 tOO al0- 20 20- 30- 40- 50 60- O— 0 200 400 600 800 _ _ ~ f -~ 40— 50- 60 FIGURE 9.9 Arsenic concentrations in sediments of lakes near Sudbury, Ontario (from Nriagu 1983). them were not initiated to study atmospheric fluxes of metals. However, there is a striking uniformity in the finding of increased concentrations of metals in modern sediments reached by a variety of researchers using a variety of techniques on a wide variety of lakes. These conclusions are consistent with data from a restricted class of lakes that over the past 10 years has been studied specifically to address the question of atmo- spheric deposition of acids and metals and associated biological effects. These lakes are oligotrophic, low-alkalinity lakes with a history of relatively stable land use and with watersheds that have been continuously forested and undeveloped. Many of the lakes are at a relatively high elevation in the northeastern part of the United States. Researchers study this type of lake assuming that changes in sediment chemistry could be assigned unambiguously to changes in atmospheric chemistry. One of the earliest studies of heavy metals in sediment cores from remote lakes was made by Iskandar and Keeney (1974) in Wisconsin. The metal concentrations in sediment were based on the technique of nitric acid extraction of the sediment, not total digestion. Concen- trations of extractable lead and zinc increased

377 Pb (ppm) 0 50 100 150 170 0 50 100 to, ~ 1932 20 20 - I 40 cat us c, 60 80 Monona Wingra 3~~ — Mendota (a) Zn (ppm) 1932 20 ~ ^ 1894 ~ 'a I 40 LL 1 856 1818 60 _ - 150 Or' .,.N ~ Minocqua Butternut Phillips (b) FIGURE 9.10 Vertical distributions of (a) lead and (b) zinc in selected Wisconsin lake sediments (from Iskandar and Keeney 1974). younger sediments in all lakes (Figure 9.10) regardless of watershed activities, including some pristine lakes in northern Wisconsin. Iskandar and Keeney postulated that lead was delivered through atmospheric processes but did not attempt to explain the elevated zinc concentrations. Norton et al. (1978), using sediment chemistry of pollen- dated cores from New England lakes ranging from culturally eutrophic to pristine and oligotrophic, suggested that the ubiquitous rise in the concentration of zinc in lake sediments in the 1800s could best be explained as having been caused by atmospheric deposition. GallowaY and Likens (1979) reported the detailed sediment chemistry of a single core from Woodhull Lake in the Adirondack Mountains of New York. The coring technique, the analytical methods used, and poor dating Prevented Quantification of sediment deposition rates. Nonetheless, increasing concentrations of many elements associated with fossil fuel burning, such as silver, gold, cadmium, chromium, copper, lead, antimony, vanadium, and zinc, led them to conclude that the increases were caused by increased atmospheric deposition of these metals. Heit et al. (1981) reported on two lake sediment cores from the same region. Although both cesium-137 and lead-210 dating were done, information from the latter method was not utilized. The cesium-137 chronology suggested that PAHs and fossil-fuel-related

378 elements started increasing about 1948, in agreement with the estimate of Galloway and Likens, also based on cesium-137. This dating methodology is now believed to be inaccurate (Davis et al. 1984), giving ages of sediment intervals that are too young and consequently deposition rates that are too high. This was the first serious attempt at evaluating atmospheric deposition rates of trace metals from analysis of lake sediments. At about the same time, in Quebec, Ouellet and Jones (1983), also utilizing cesium-137 for dating, came to conclusions similar to those reached by Galloway and Likens as well as by Heit et al. with respect to the timing of the increases of heavy metal loading (Figure 9.11). This consensus "demonstrated" a regional and apparently synchronous air pollution phenomenon. Perhaps the best way to date sediments is to use varves or varves in conjunction with other dating techniques. Johnston et al. (1982), using varves, pollen, cesium-137, and lead-210, concluded that for three lakes in Maine accelerated sedimentation of lead and zinc started well before 1950, perhaps as early as the late 1880s. Tolonen and Jaakkola (1983) have nearly perfected these verve-dating techniques, studying varved sediments in Finland and calculating deposition rates for metals with great precision. I. Renberg (Umea University, Sweden, personal communication) has been able to evaluate sediment strata year by year using this approach. Unfortunately, lakes containing varved sediment are rare in the United States, and few have been well studied. Norton et al. (1982) and Kahl et al. (1984), studying a variety of lakes in northern New England as well as in the Adirondack Mountains (Norton 1984), have utilized lead-210 chronology for sediment dating and have concluded that the increase in heavy metal loading, as indicated by increasing concentrations of lead, zinc, and copper, across the New York-New England region, started in the period 1850 to 1900; Evans et al. (1983) and Evans and Dillon (1982) reached the same conclusion for Ontario. Starting about 1930 to 1940, the concentration of vanadium increased concurrently with the rise of consumption of oil, much of which is vanadium rich. Deposition rates for elements that are particularly enriched (e.g., lead), increase as well (Figures 9.12, 9.13). The concurrence of the increase in deposition rates of lead, zinc, and vanadium in a variety of lakes in the northeastern United States and eastern Canada

379 - - c' CD ID rot _ N ~ ~ g ~ q 8 _ oO _ _ X _ _ 11 11 DATE .~ co l 20 40 60 80 DEPTH (cm) 1.60 _ a' 1.20 _ ~ 0.80 _ 0.40 _ 0.00 G E E E Q Q CD a, I {:L CL a LO 1O DATE DEPTH (cm) FIGURE 9.11 Sediment chemistry from lakes Tantare and Laflamme, Quebec, Canada (from Ouellet and Jones 1983). SAEF, sedimentary anthropogenic enrichment factor; pCi/g, picocuries per gram; ppm, parts per million. The value of SAEF is obtained by dividing the measured concentration by the background concentration. Thus, because the background value of picocuries per gram of cesium-137 is negligibly small, the SAEF value was assigned a value of infinity (a).

380 2000 1900 1 Ann ~ Inns Anne 1 Ann ~ 2000 - Pb-210 TiO_ 1900 f 1 800 - 1700- 1600- 1 500- ... .. .... 0 50 pCi/g DRY 0 100 PCT. WET 2000 Na2O 2000 K2O 2000 CaO 1900~ 1900 ~ 1900 _ 1800 _ 1800~ 1800— 1700i it 17oo 1_ ~ 17 1600 1 ~ 1600 1 ~ 1600-I ~ 1 500 1 ~ 1 500 L 1 500 1 ~ · 60 o.o 1,0 0.0 1.5 v.v PCT. IGN. PCT. IGN. PCT. IGN. ~ l 1900.1 1 800 1700. :'onn MnO ` ~ =t 1600 \ 1 500 ., .,,,, ,, 0.0 0.1 _ PCT. IGN. PPM IGN. 2000 1 900 - 1800- 170 1 600 - ~ 1500 . 0 700 _ A12O3 2000, -Z 2000 1, ~ _ 1 900 ~ 1 900 ~ 1800 at_ 1800 ma 1700 ~ `', 1700 1600 ~ ' 1600 1500 1 ~ , . 1500 1 ~ o.o 0.6 0 20 PCT. IGN. PCT. IGN. 2000 MgO it_ 1 900 i 1 800 ~ 1700 1 600 1500 1.5 0.0 PCT. IGN. 2000- On 2000 - C_ 2000 _ 1900 r 1900- ~ 1900 i 1800 ~ 1800 _: 1800 _ 1700 = 1700 ~ 1700 1 600 . ~ 1 600 - ~ 1 600 ~ ~ 1 500 , ,, ,, , 1 500 , .,, ,, 1 500 1, O 1 000 0 ; 0.0 2000 FeO _ 1900 1800 1700~ 1600 - . 1 500 1 · · ~ 1.0 0 20 PCT. IGN. 70 0 200 , PPM IGN. PPM IGN. PPM IGN. CONCENTRATION VS. YEAR FIGURE 9.12 Sediment chemistry from Jerseyfield Lake, Adirondack Mountains, New York. PPM ION., parts per million, ignition; PCT. ION., percent, ignition; pCi/g DRY, picocuries per gram, dry; PCT. DRY, percent dry; PCT. WET, percent wet. (From Kahl and Norton 1983.) suggests that atmospheric deposition, among other factors, is the source of the metals. Evans et al. (1983), Dillon and Evans (1982), and Evans and Dillon (1982) studied the chemistry of sediments from 10 lakes. They used multiple cores, dated them with lead-210, and assessed total lake burdens for zinc, cadmium, and lead. (Coring and analytical techniques differed slightly from those of Norton et al. (1982).) There was no relationship between watershed size and metal burden. Although concentrations of trace metals in sediment varied appreciably with water depth, where the cores were taken (Figure 9.14) in the various lakes the whole-lake burdens of these metals varied only

381 4.0 3.0 2.5 UJ a: z o o CL LU 2.0 1.0 0.5 Pb JERSEYFIELD 0.0 2000 1 900 b sn ~ - 1, E ~ 3.0 LU z - ~n o CL 1800 1700 1600 1500 CALENDAR YEAR 2.0 I,, ,~ JERSEYFIELD 0.0 2000 1900 1800 1700 1600 1500 CALENDAR YEAR c 1.1 1 .0 - ~ 0.9 en 3 0.8 z o - ° 0.6 0.7 0.5 1 ~ 0.4 2000 1900 1800 1700 If v JERSEYFIELD if 1600 1500 CALENDAR YEAR FIGURE 9.13 Calculated deposition rates for lead, zinc, and vanadium over time from Jerseyfield Lake, Adirondack Mountains, New York (from Norton in press). over a range of about 100 percent, suggesting significant atmospheric input and retention of the atmospherically derived metals. Of these metals lead is retained the most efficiently, zinc the least efficiently. Evans and Dillon (1982) reconstructed the approximate atmospheric loading for lead, based on the multiple-coring approach,

382 250 v, c 200 G. 1 50 z o , - 100 z IL c' z N 50 o 500 v, E :~ 300 o '_ 200 z z o c) N 400 o \ HENEY _ , \ , ~ ~ , ~ ~ 3.9 m ___ 4.3 m .. ~ 5.5 m . 4.0 m ..... 4.5 m , 1 1 1 1 0.0 0.5 1 .0 1.5 CDMA (g/cm2) -~! ,, ~\W ; . ,~ . 1 0.0 0.5 2.0 2.5 CLEAR 0.0 m _ - - 21.6 m . 15.1 m ..... 26.1 m 1 1 1 1 1.0 1.5 2.0 2.5 CDMA (g/cm2) PLAST I C c 300 - z O 200 ~: z z o 100 N o , '''N'\\~ L 0.0 0.5 ~ , _ . 9.8 m ___ 11.9 m _. 10.1 m ..... 15.8 m . 17.5m 1 1 1 1 1.0 1.5 2.0 2.5 CDMA (g/cm2) FIGURE 9.14 Zinc concentrations versus CDMA (cumulative dry mass (of sediment) per unit area) (from Evans et al. 1983). CDMA, an alternative way to express depth profile, is employed to take into account sediment compaction. Thus g/cm2 corresponds to sediment depth; e.g., the lower the value of CDMA, the more recent the sediment.

383 using whole-lake lead burdens through time. The chro- nology of the increasing concentrations and deposition rates is similar to that reported for remote lakes in New England and New York. For example, Tables 9.1 to 9.3 list data for copper, lead, and zinc, respectively, showing enrichment factors as well as increased accumula- tion rates for 10 lakes in the Adirondack Mountains of New York. In spite of widely differing limnological and watershed characteristics, the sediment records give comparable conclusions. This selection of representative data from the remote lakes of eastern North America demonstrates the influence of an improving analytical technology on the evolution in our understanding of the record of changing atmospheric deposition of heavy metals as revealed in lake sediments. Neither the concentration nor the deposition rates of the heavy metals can be directly related to the acidity of precipitation, however. Elevated concentrations of certain trace metals in sediment only indicate the existence of polluted air masses. The acidity of pre- cipitation in the eastern part of North America is due predominantly to the presence of sulfuric acid in precipitation (National Atmospheric Deposition Program 1984) with geographically and seasonally variable contributions of nitric acid. The proportional con- tribution of nitric acid to the acidity of precipitation has probably increased in recent decades (see Chapter 2), while regional sulfur emissions in the northeastern United States have probably not changed substantially in the past 50 years (see Chapter 2). Several studies (e.g., Mitchell et al. 1981, 1985) have attempted to relate the distribution of sulfur in sediments to atmo- spheric deposition of sulfate. Sulfur may reach the sediment through a variety of pathways, including sedimentation of allochthonous and autochthonous organic matter, adsorption of sulfate on sedimenting detritus, sulfate reduction with precipitation and sedimentation of sulfides from the water column (Mayer et al. 1982, Schindler et al. 1980), and sulfate reduction within the sediment and precipitation of sulfides at depth within the sediment (Norton et al. 1978, Car ignan and Tessier 1985, Holdren et al. 1984). Thus the sulfur content of a particular interval of sediment, and even the speciation of sulfur in that interval, may bear no direct relation- ship to atmospheric depositional values. Sulfur profiles in sediment vary widely from lake to lake (Figure 9.15). although the general history of sulfur deposition is

384 3 By U: ._ Ct o C) Ct A: sit .~ ._ U) on U. o V I: ._ ._ o ._ o V o V Ch m 0~ 1= To 3 - 0 O ·— ~ ._ a_ Ct ~ ~ m as' 0 Ct c I E .~o ;> ~ it' C m _ ~ To ~ ~ 0 — ~ ~ ~ m 0 Ct E E C ~ to o c=5 m C) to ~ t——0 ~ 0 ~ 00 ~ . . . . . . . . . . Cal ax ~ 0 0 — - , — ~ ~ 00 00 00 00 00 00 ax ~ ~ at ~ ~ _________ ~ GN ~ ~— - 1 00 ON — 00 ~ . . . . . . . . . . O O O O O O O O O O O ~ O ~I t— —~ 3N ~ ~ _ ~ _ O _ ~ O — O . . . . . . . . . . O O O O O O O O O O ~1 ~ ~ '= O ~ C~ ~ t— . . . . . . . . . . ~ ~ ~ ~ ~ ~ ~ _ ~ C~ a~ 00 00 00 ~ 00 ~ C~ ~ ~ ~ C~ ~ 0- _ _ ~ _ _ _ _ ~ _ _ - ) ~ t— — ~ — ~ 00 ~ a~ ~ ~ ~ ~ ~ ~ oo _ ~ ~ ~ O _ O ~ 00 - , _ C) Ct ~, ~ E —E C~ o ._ - a~ - Ct a~ ._ o o ._ ~_ .s ~s Ct Ct U. ._ - c ._ Cd o ._ s~ oZ b4 C;S <5 ~ - o ~r oo p~ ~0 ~ ~n

385 50 >lo, By Cal 'Ct o C) Ct s°- ._ a ._ U) Cal of Cal so o V ._ V) ._ o . _ o V Cal m O = o m O Ct =~ .4 O ;. ~ it m O O O A ~ m O en C _ Ct ~ ~ O ~ =: Ct Pa ~ 3 V) ~ ~ ~ ~ ~ ~ _ Go ~ ~ ~ o ~ ~ o to ~ ~ ~ Do Do Do Do ~ ~ ~ cr ~ a~ ~ ax a~ ~ ~ ~ ~ ~ ~ ~ ~ ~ _ ~o ~ ~ ~ o ~ o ~ ~o 0 0 ~ ~ r~ u~ oo ~ ~ ~ . . . . . . . . . . r~ ~ ~ ~ cr~ oo - ~ o . . . . . . . . . . o o o o o o o o o o - ~ oo ~ ~ ~ ~ c~ oo ~ ~ ~ - ~ ~ - - ~ ~ ~ ~ a~ 0 ~ ~ oo c~ oo ~ ~ ~ o c~ ~ ~ a~ ~ ~ ~ ~ ox - - - - - ~ - - - - oo o o ~ ~ oo ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ c~ ~ ~ ~ u~ ~ o o ~ ~ ~ ~ ~ ~ - ~ ~ ~ ~ ~ ~ ~ c~ ~ ~ y ~ ~ ~: o . - au o - e :: so-. . - o & & c~ o & . . u' e c~ . - & . - o . - u, 50 . - o o z ~4 e . . o ~ 00 ~ 1 - ~ _ ~ O ~ C~

386 JO 3 By U) · Ct o hi: C) so ._ ._ U) en on o U. o V A ._ V) A: ._ o ._ Ct o V ._ m ~0 '= 00 ~ O .~0 ~ ~ m ~0 Ct ~ .E = o >1 . o m 0 ° O ~ ~ o Ct ._ o Ct m _` S:~ o C5 C) o C) a: ~ ~ ~ ~ ~ Do ~ ~ ~ ret 00 ~ ~ 0 0 ~ ~ ~ 0 00 00 ~ to at _________ o ~ ~ ~ ~ ~ ~ ~ o ~ ~ ~ ~ o oo ~ ~ ~ ~ . . . . . . . . . . ~ ~ ~ ~ ~ ~ o ~ ~ ~ - ~ ~ oo oo ~ ~ oo oo ~ ~ . . . . . . . . . . o o o o o o ~ o o o ~ ~ ~ ~ ~ ~ ~ - ~o o tn a~ - ~ ~ - - - ~ - - - _ ~ ~ ~o oo ~ oO In ~ oo cr~ ~ —~ ~o ~ ~ ~ 0 - ~o~ooooo~o ~ ~ ~ ~ o u~ ~ ~ o ~ - - - - - ~ ~ - - _4 ~ ~ ~ ~ E ~ o ¢ o ._ C) o P~ - ~o ._ C) o 3 C~ o s~ ._ ._ U: Ct Ct ~: C~ ._ - rD ~: .= o .~ C~ .> ~: o oZ ~ . =¢ ~' . . O ~ ~ C ) oo~ O~ ~0 ~ cn

387 . Sit E 10 Q ~ 20 . . . IS ?~ _ . . LEDGE PON D . ,umol/g (dry mass) 0 100 200 . O 300 4 00 1 . 30 1944 1846 In O 1712 ~ 1 598 1-LJ ,umol/g (dry mass) o 1 00 200 300 4 00 or At E 10 I_ I 15 lo 20 2S 30 . . . . . SOUTH LAKE FIGURE 9.15 Sulfur concentrations (micromols of sulfur per gram of dry mass) in sediments of Ledge Pond, Maine, and South Lake, New York (from Mitchell et al. 1985). 1957 C) LO An O 1859 Q C) 1801 <t LL similar. Concentrations of total sulfur in sediment do not appear to reflect the atmospheric deposition history as inferred from sulfur emission estimates (see Chapter 2). Studies of the isotopic composition of sulfur in cores of sediment have been attempted to ascertain the history of atmospheric deposition of anthropogenic sulfur (Nriagu and Harvey 1978). This has been done largely to cor- roborate the chronology of deposition of other con- stituents, such as lead, and also to shed light on terrestrial and aquatic processing of sulfur. The studies are in their infancy and do not yet permit a clear understanding of processing. Shiramata et al. (1980) have studied lead-isotope ratios to discriminate between anthropogenic atmospheric lead and lead derived from the drainage basin. These studies are also not yet sufficiently developed to be useful in studying trends in deposition. A number of solid products are formed and injected into the atmosphere as a result of the combustion of fossil fuels. Griffin and Goldberg (1981) demonstrated a vari- able trend in the concentration of charcoal particles in the sediment of southern Lake Michigan. An increase in concentration began about 1900, peaked around 1960, and has declined slightly since then, possibly due to increased retention of fly ash at the sources of com- bustion (Figure 9.16). They interpret this trend as a reflection of emission rates for particulate matter in

388 0~ 1970 1960 1950 1940 1 930 1920 1910 1900 1890 1880 1870 1860 1850 1840 1830 10~ 20 _ - I UJ _ 30 _ 40 _ 50 _ 60 o ~ T I I I I I ~ I I I I T I I I An1 I I i i i r- I ---I v 1 ~ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 020 % CARBON by WE IGHT >38 010 FIGURE 9.16 The record of concentration of charcoal particles in a sediment core from southern Lake Michigan. Particle sizes are greater than 38 microns (from Griffin and Goldberg 1981). that region. They also note that particles from coal combustion have been dominant since 1900. However, particles derived from combustion of oil first appear around 1950 and have increased since then. Renberg and Wik (1984) have found similar trends and used their results to develop a chronology, using soot particles, in varved lake sediments in Sweden. The soot chronology is so ubiquitous in Swedish lakes that it can be used as an independent chronological tool. Concurrent with the increase in fossil-fuel-related charcoal is the concen- tration of magnetic spherules produced by high-temperature combustion.

389 PAHs in sediments may be produced by natural processes or enhanced by bioaccumulation of trace amounts produced by other natural processes, such as diagenesis of organic matter and natural fires (Base and Hites 1976). However, it is clear from analysis of sediments predating the industrial revolution that concentrations of natural PAHs in sediment are very low, and certain PAHs are known to be produced only by fossil fuel consumption. Numerous investigators have detected high levels of anthropo- genically produced PAHs in surface sediments and cores from polluted lakes, but the sources of the PAHs have not been determined unequivocally (e.g., Lakes Lucerne, Zurich, and Greifensee in Switzerland; Wakeham et al. (1980)). Heit et al. (1981) and Tan and Heit (1981) reported biogenic and anthropogenic PAHs in sediment cores from two relatively pristine lakes in the Adirondack Mountains of New York. Although the chronologies of these cores are not precisely known, the authors demonstrated dramatic downcore changes in PAH concentrations. Correlations with atmospherically derived regional indicators of atmospheric pollution, such as lead and zinc concentra- tions, are positive, so the trends in the PAH concentra- tions are likely to represent trends in regional emissions. Several other studies report comparable results from a variety of lakes in the Northeast. R. Hites (Indiana University, personal communication) is now evaluating PAM data from well-dated cores in New York and New England for which ancillary data, on heavy metals, for example, are available (Figure 9.17). There seems to be little doubt about the general time relationships between PAH abundance and other indicators of polluted precipitation, particularly the onset of pollution. M. Heit (Department of Energy, Environmental Measurements Laboratory, New York, personal communication) has suggested that the decrease in PAH deposition rate over the past 20 to 30 years (see Figure 9.17) may be related to a reduction in low-temperature burning of coal, specifically from home heating. Although profiles of metals and other pollutants do not reflect directly the acidity of precipitation, sediment chemistry is influenced by lake-water pH. This influence has been studied experimentally and empirically. Acidification of laboratory microcosms (e.g., Hongve 1978, Norton 1983, Kahl and Norton 1983, Norton and Wright 1984) and experimental whole-lake acidification

390 7 6 5 4 N 3 2 o 1800 1 820 1840 1860 1880 1900 1 920 1940 1960 1 982 210Pb DATE Pb I_/ _ — 1 1 1 1 1 300 200 1 Cat 100 PAH TV in/ \_\ OL I ~ I I I I I I 1 800 1820 1840 1860 1880 1 900 1 920 1 940 1 960 1 982 210Pb DATE FIGURE 9 .17 PAH and lead deposition rates for a core from Big Moose Lake, New York (lead data from Norton, unpublished; PAH data, courtesy of R. Hites, Indiana University, Bloomington, unpublished). Unit of mass given for PAH deposition is nanogram (ng) equal to one- billionth of a gram.

391 (Schindler 1980) suggest that several elements, at least zinc, manganese, and calcium, are lost preferentially from sediment as pH is lowered. Kahl and Norton (1983) were able to deplete substantially the tops of sediment cores of zinc, calcium, and manganese during experimental acidification of microcosms. Other elements, such as aluminum and iron, were released in substantial amounts, but the reservoir of these elements in the sediment is sufficiently large so as not to be substantially altered by changes in the pH of the overlying water. Dickson (1980), Norton et al. (1980), and Evans et al. (1983) arrived independently at the conclusion that zinc concentrations were lower in recent sediments in lakes with low pH (<5.5). Concentrations of calcium and manganese also seem to be lower in recently deposited sediment in acidified lake sediment cores (see Figure 9.12). These declines start in sediment as old as 100 years. It is not clear whether the observed reductions occurred as a result of postdepositional processes causing accelerated rates of release of metals from the sediment, a synchronous process causing leaching of metals from sedimenting particles, or predepositional conditions causing reduced levels of metals to deposit on sedimenting particles. Carignan and Nriagu (1985) have observed active loss of manganese from sediments, in situ, where the overlying water had been acidified by acidic pre- cipitation. If the loss of metal is in situ it could affect sediment deposited as much as 10 to 20 years older. If the loss of metal is predepositional and not caused by post-depositional diagenesis of sediment, trends in zinc, calcium, and manganese sediment deposition rates, if they are independent, may be useful in tracking the process of lake acidification. Peaks in concentration and apparent deposition rates of zinc below the surface sediment in acidic lakes may be related partially to sulfate reduction in the sediment and precipitation of zinc sulfide below the sediment/ water interface, i.e., in older sediment. This effect has been observed by Carignan and Tessier (1985) in Clear Lake, near Sudbury, Ontario, but the general occurrence of this phenomenon has not been demonstrated; it cannot explain calcium and manganese depletion in sediment nor reduction of the concentration of zinc to values lower than background values. This has been observed in several lakes with pH below 5. In summary, multiple lines of chemical evidence derived from lake sediment indicate a history of changing

392 atmospheric chemistry over a time frame of about 100 years. Sediment constituents related to coal burning and smelting, which include soot spheres, magnetic spherules, PAHs, charcoal, metals, and excess sulfur, increase in concentration dramatically through the first half of the twentieth century. Pollutants associated with combustion of fuel oil (glassy spherules and excess vanadium) and gasoline (lead) increase over the past 40 to 50 years. The synchronous rise of these constituents in sediments in both remote and urban lake sites in the northeastern United States and eastern Canada implicates atmospheric pollution as the likely source. Evidence for the actual acidity of precipitation cannot be inferred from the chemistry of sediments but decreases in the pH of aquatic systems are suggested by recent declines (relative to immobile elements) in the deposition rates of elements mobilized by increasing lake-water acidity. The elements include zinc, manganese, and calcium. THE CHEMICAL STRATIGRAPHY OF PEAT BOG DEPOSITION Ombrotrophic bogs, by definition, receive all their nutrients from the atmosphere. Sphagnum, a dominant genus in these bogs, has a very high capacity for cation exchange and adsorbs and incorporates considerable quantities of metals, releasing hydrogen ions (acidity) in the process. Consequently, many researchers have utilized the chemistry of surface peats or peat land plants to establish a regional pattern of atmospheric deposition of pollutants. For example, Pakarinen and Tolonen (1976) and Steinnes (1977) demonstrated regional trends in heavy metal concentrations in lichens and mosses for Finland and Norway, respectively. They concluded that spatial concentration gradients were related to emission gradients. Data from Groet (1976) indicate high concentrations of metals in mosses in the northeastern United States; regional emissions of metals are believed to be responsible for the high concentra- tions of metals. Neiboer et al. (1972) give evidence for a strong relationship between concentrations of heavy metals in lichens and metal emissions from the nickel smelter at Sudbury, Ontario. Thus the spatial chemical pattern displayed by modern lichens and mosses may be used to demonstrate snort- and long-range transport of metal pollutants.

393 The precise relationships between depositional flux of various metals and their retention by the contemporary moss or lichen surface are not known. Atmospheric con- centrations, precipitation amount and type, and plant species community composition are probably all important factors in the relationships. Analysis of data in the literature for Scandinavia (Pakarinen 1981a) suggests that over the decades of the 1960s and 1970s, metal concentrations increased in mosses/lichens, implying that atmospheric deposition rates increased, too. Pakarinen (1981a) repeated measurements of contemporary Sphagnum in Finland from 1975-1976 to 1979-1980 with the same analytical protocol and found significantly declining concentrations of lead, zinc, and iron; copper was unchanged. These declines occurred in a time of reduced emissions and declining use of lead in gasoline in northern Europe. This evidence suggests that trends in atmospheric deposition of metals may be monitored using mosses or lichens. However, total retention of metals after initial sorption by the living surface has not been demonstrated and for most elements is probably unlikely. Pakarinen (1981b) has demonstrated that the order of decreasing mobility of elements in a recently formed Cladonia arbuscula profile is potassium ~ nitrogen > phosphorus ~ magnesium > zinc > copper > manganese > calcium > lead > iron. Damman (1978), in a classic study, demonstrated that in a Sphagnum-dominated hummock profile the order of decreasing mobility is potassium > sodium > lead > phosphorus > nitrogen > magnesium. It is clear that the chemistry of cores of accumulating ombrotrophic bogs do not accurately portray atmospheric deposition chemistry nor the chemistry of the initially formed organic matter. These conclusions were anticipated by the experiments of Clymo (1967), who showed that aerobic decay of at least Sphagnum proceeds at considerable and measureable rates, releasing nutrients associated with the biomass. Mosses (especially Sphagnum) do not obtain all their nutrients directly from the atmosphere. Much of the nutrient pool is recycled, and because of decay the chemistry of these mosses is not fixed after death. Therefore, reconstructions of trends in atmospheric deposition are difficult to establish from profiles of chemistry. The best opportunities for reconstructions exist where the living surface is just at or only slightly above the water table (i.e., a bog hollow) . . . .

394 BIG HEATH CORE 3 (HOLLOW) o- 10- 30 - 40 - 80 100 PCT. WET Na n- 10- 20 - 30- - 40- ORGANIC H2O 0 CARBON Pb-210 10~ 10 ~ 30 l l 40- l l l l l l l l l _ l l l l l l l l l 0 20 pCi/g DRY Ca 0 800 PPM DRY 0T 10 20 - 30 - 40 - 20- 30 - 40 90 1 00 PCT. DRY K 0~ 1 0- 20 30 - 40 ......... 0 500 PPM DRY O- 10- 20 - 30- 4'0- Pb Zn ~ 1°: i7 20~, 0 1 00 0 50 PPM DRY PPM DRY Fe 0= 20- 30- ~ ~ TP 1 40 ~ j ~ l l l l l l l l l 0 1500 PPM DRY Mg 20 ~ 30 - _ ......... 0 5000 PPM DRY Mn O. · 1 r-.l I 10 t ~ 1 G I 20: 30 ~ 1 40~ 1 l l l l l l l l l l l 0 50 PPM DRY Al 40 0 2000 PPM DRY 0q 10] 20] 303 40~ ,,,, o - 10 20 30 40 - 1or llill 20 - 30= 40t—, , . , . . 1 0.0 5.0 0 1 00 PPM DRY PPM DRY CONCENTRATION VERSUS DEPTH (cm) 1 .... ~ ..,, 1 0 2000 PPM DRY FIGURE 9.18 Chemistry of a hollow peat core from Big Heath, Mount Desert Island, Maine. Water table at about 5 cm. PPM DRY, parts per million, dry; pCi/g DRY, picocuries per gram, dry; PCT. DRY, percent dry; PCT. WET, percent wet. (From Norton 1983.) There, the older moss is below the water table where decay and chemical alteration are retarded. Damman (1978) showed that a hollow core in Norway yielded a lead profile that resembled those derived from lake sediments (see, e.g., Norton and Hess 1980). In a similar study In Maine Norton (1983) (Figure 9.18) found nearly identical results. However, comparison of hummock and hollow peat profiles suggests that multiple processes are affecting their chemistries (see Figures 9.18 and 9.19). These processes probably include the following as the most important: upward and downward biological pumping, involving, for example, cesium-137, potassium, and manganese; downward or upward migration of mobile elements during precipitation events or capillary

395 BIG HEATH CORE 1 (HUMMOCK) 2O O- ~ 10~ 20- j 30- 40- 1 50 ~ ........ . 80 1 00 PCT. WET Na 0- 10 ~ . 20 ~ 30~ . 40 ~ 50 - l ORGAN I C CARBON A Pb-210 Fe Mn ~ 10c 10~ 30~ '1 30i,,,~ 50 I,...... O- ~ ., O- 10 - l 10 20 - 1 20- 7 30 ~ ~ 30 r 40 ~ . 40 50 - , so . ~- 90 100 PCT. DRY K O 500 PPM DRY o 10 20 30 40 50 Pb . U....1 0 150 PPM DRY u- ~ 1 10 - 20 ~ ~ , 30 - 40' ~ 50 ~ ~ . I, 111 1 1 1 1 O1 10] 20g 305 40~ 50q . (, 1 0 15 0 2500 pCi/g DRY PPM DRY Ca Mg 0 150 PPM DRY Al . . ~ . . . O 1 500 0 5000 0 2000 PPM DRY PPM DRY PPM DRY O- 10 - 20 - . 30 ~ . 40 ~ 50 - 0 100 PPM DRY u 10 20 30 40 50 ~ 1 ....... 0.0 8.0 PPM DRY CONCENTRATION VE RSUS DEPTH (cm) O- . 10 - . 20 ~ 30 - . 40 - 50- . ~ ' 0 100 PPM DRY 0 2000 PPM DRY FIGURE 9.19 Chemistry of a hummock peat core from Big Heath, Mount Desert Island, Maine. Water table at about 30 cm. See caption, Figure 9.18 for definitions. (From S. A. Norton, University of Maine, Orono, unpublished data.) winking; concentration of immobile elements, such as iron, manganese, and titanium by biological decay and selective removal of mobile elements; and redox reactions related to the water table affecting, for example, zinc, iron, and possibly manganese. The profiles shown in Figures 9.18 and 9.19 are similar to those given by Damman (1978) and probably represent end members for chemical profile development. Intermediate water table conditions probably produce intermediate-type profiles, and the presence of certain vascular plants drastically alters the mobility and distribution of some metals. For example, high concentrations of surface manganese are associated with shrubs. In their absence, manganese in surface peats is low in concentration.

396 20.0 18.0 16.0 14.0 12.0 in o 10.0 - - in C: 8.0 6.0 4.0 2.0 o 1 1 1 1 1 1 1 60 50 40 30 20 DEPTH IN CENTIMETERS (a) 10 0 FIGURE 9.20 (a) Cesium-137 (Cs-137) activity in a hummock peat profile, Bull Hummock, in Bull Pasture Plain, New Brunswick (August 29, 1981) (from Olson 1983); (b) annual deposition of strontium-90 (Sr-90) in New York City. Cs-137 deposition = 1.7 x Sr-90 deposition. (From Larsen 1984.)

397 30.0 25.0 o ~ 20.0 I 15.0 10.0 5.0 o FIGURE 9.20 (continued) . ..... , , . . , . ; , , .... E :....: :.:.:.: :.::::: :::,:, ,:.:.:.::::::::::::. E: ::,.,.:':.: .:,:,:,:, :, I,:.:.:: .,..,,. :.,:, ,: 2.: ... : : ::::::::,:::::::::.:::::::::::::::::::: : :::: :::.::: ::::::::::: :.:.:.:.: :.:.:.::::::: ::::::::::::: ....—................................................ :::::::: :: :::::::::::::::::::::: :::::::::::: ~ ..,: ... i. i:..:: :.. ,..:: :: 2. :: :' ., .. :, .............................. ............. ~2'2'''.''.".2.''.2.""""""""""""""""""""""'"""~ :::::: ........ ::.::::: ::: :: ............ :::::,::: .. . ....'........ .'........ .,....:,...... .......... . ::::: . ........ ....'', .,.,.,.—,...... ::.:.:.::::::.:::::: :.: :. : :.:,::::::::, : :.:.:.:.:.:.:.:.:.: .......................... .::::: :.:.:.:::.:.:::.:::.: :,::::: :::::::,::::: ,—,..,., .'..."..'.................. :,. :,:.:: :.: ::::::: :, :.:::.: .. 2 : ',.: 2 " 2. :: ,: :,,:::::,:,:: ,: :::::::.::::::::::: ....................................... :,:,:,:,:,:,:,:,:,:.:,:,:,:: :,::::: :.: . j .............................................. .............................................. : :...:. . .:.: :.. .: :: . ~ .~ ........................................................ :::~::: :::::.::.: ::::::::::::::::::! ::: ::::::::] 1955 1 960 1965 1 970 1975 1980 YEAR (b) The development of a reliable dating tool for peat is an additional problem in utilizing the chemistry of peat profiles for establishing chronological trends in atmo- spheric deposition of pollutants. Dating methods include pollen stratigraphy; chronostratigraphic markers, such as charcoal, cesium-137, and lead-210; increment dating; and magnetic stratigraphy. Pollen stratigraphy and the use of charcoal are typically imprecise for dating purposes because atmospheric particulate material can penetrate through the living surface into older peat. Continuous age-depth relationships cannot be established; only one or several time lines are established in the stratigraphic section, and extrapolation and interpolation of ages are necessary. (See Livett et al. 1979.) This shortcoming also arises with the use of cesium-137, a product of thermonuclear weapons testing, as a stratigraphic marker It gives one (the 1963 maximum) or at best two ages. (See Clymo 1978.) Also, it has been well established that cesium as well as potassium is strongly recycled by .

398 or into living moss, and pumped downward into older peat by vascular plants (see Oldfield et al. 1979). The resulting cesium-137 profiles may bear little resemblance to the known deposition history of cesium-137 (Figure 9.20). During the past 10 years, several workers have used lead-210, a naturally occurring radionuclide deposited from the atmosphere, to establish the chronology of cores and thereby interpret trends in atmospheric deposition of certain constituents. Lead-210 dating assumes no penetration of this element through the living surface during initial deposition and no vertical displacement of it from the stratum in which it is deposited. Views on the conservative and immobile nature of lead differ widely. One of the first uses of lead-210 for the dating of peat in the United States was by Hemond (1980). Although the stratigraphic resolution of the peat profile from Thoreau's Bog, Massachusetts, was poor for purposes of trend analysis (four increments of peat for the past 100 years), the derived atmospheric loading rates of lead are comparable with values derived from lake sediments (1971-1977, 43 mg m_2 yr~l; 1948-1971, 44; 1920-1948, 33; 1881-1920, 11; 1844-1881, 12). Increment dating of peat (Pakarinen and Tolonen 1977) is based on counting annual nodes in moss species. Comparison between increment dating and lead-210 dating suggests that in hummocks, lead-210 may be mobile with downward dislocation of lead. Lead-210 analysis gives ages that are older than increment ages and are thus too old if the increment dating method is correct (Figure 9.21). This effect would artificially elevate deposition rates based on lead-210 in recent years. Increment dating cannot be utilized in hollows, where biological decay apparently destroys the evidence. On the other hand, pollen dating and lead-210 chronology appear to agree for hollow profiles (Tolonen et al. in press, Norton in press). At best, it appears that dating of peat profiles is somewhat imprecise. Consequently, the chronology of trends in peat chemistry, and thus inferred changes in atmospheric deposition of metals as well as net deposition rates, are imprecise. Madsen (1981) showed increasing levels of mercury over the past 200 years in lead-210-dated cores from two Danish bogs and suggested that this trend in mercury concentration relates to increased emissions from industrial or agricultural activity. Pakarinen et al (1980) used increment dating to establish net loading

399 100 . 111 _ · / =50 _ · ~ · / _ . / O / 1 1 1 1 1 1 1 1 1 1 0 50 100 210Pb AGE FIGURE 9.21 Increment age versus lead-210 age for a hummock peat core from Great Heath, Cherryfield, Maine (from Norton 1984). rates of a variety of metals including lead, zinc, and copper, over the past 10 to 20 years. The values are similar to values of measured atmospheric deposition. However, strong redistribution of zinc and iron appears to occur. S. Norton (unpublished data) has dated, by lead-210, and chemically analyzed 10 pairs of cores from adjacent hummock-hollow associations from ombrotrophic bogs in Maine. Only cores from hollows have heavy-metal stratigraphy and deposition rates, especially for lead and zinc, that resemble those of lake sediments in the eastern United States (Figure 9.22); the resemblance is crude at best. Oldfield et al. (1978) have developed techniques for measuring the magnetic particle content of peat profiles

400 1 .0 0.8 N C) UJ ~ 0.6 By o en O 0.4 UJ 0.2 _ _ O 11 1 1 1 1 1870 1890 1910 1930 1 1 1 1 ~ 1 1950 1 970 1990 1 1 YEAR FIGURE 9.22 Deposition rate of lead for a peat core from a hollow in Crystal Bog, Maine (from Norton 1983). and found a stratigraphy mimicking reconstructed histories of emissions for the local regions near the bogs (Figure 9.23). The particles are generated as part of the fly ash in coal-based emissions. However, Oldfield et al. believe that the transport distance is relatively small . The metals lead and zinc r however, are known to be transported in large proportions for hundreds to a few thousands of kilometers. Thus, cores may give information about both short- and long-range transport of atmospheric pollutants. In summary, modern peats record spatial variation in recent atmospheric deposition of chemical and particulate pollutants; such peats typically have higher concentra- tions of pollutants than preindustrial peat. However,

401 20 o cot 40 - E Q 20 REGENT ST. BOG FREDRICTON New Brunswick 1 982 Core _ . - ~ ~ o =, 400 E O . 1000 _ 800 _ cn _ E 600 _ ~ _ LL 400 _ 200 _ n _ 3000 _ 1 000 _ _ cat _ ID to - O _ o 2000 8 So O _ _ —... ....... _ ''.'. ~ 1 1 i.... i '''' '' ....—................................. ................................................................................................................................................................................................................................................... ...................................................................................................................................................................................................................................................... ,.,.,., .,, , ,., , , ,, ,., ,.,., ,: , ~ , : 1900 1950 ~ _ i..... ............ ...'...... .......... .... ...... , ... .. ... ,..: ............. ............ ...... ........... . . ............ .... in,, .,, .,,, ., ..,,, ., ... .., .,, It c:: :.::::::,::,:,::,::,:~::~:~:::::~:::::~::::~:::~........ . ~ : i ............ ... , ,. . . . . rim ... . , .. .. l::::::: :::::::::::::::::::: ::: :.: :::.:.::::: :~::::::: ::::::::::::::::::::: :::::.:::::: ::::::::::::::::::::::: :::-:::: l: :::::::::- :: .:-:::~:::::::::::::::::~::::::::~:::.:~:::::::::-::::-:::::::: ::::::: my - - ~ : ::::: :-:::: ::::: :-::::::: :::::-:.—':::::::::::::-:-:::::::::::::::::::::::::::::::::::::::::::::::::::::::-:-:-:::-::~::::-::-:-::: : :.: : ::::::~: :::::::::~:~::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::: :::::::::::: ::::::::: :.:~::~:~:~:~:::~:~:::~:~:~:::~:::::.:-:~:~:~::.:-:~:~:~:~:~::::::~::::-::::::::::::::::::-:::::::-:::::::::-::::::::-::::-::: :::::-:::: :~:-:::-:: :: ::: :: ::::::::::::::: ::: :::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::::::::::::::::::::::::::::::::::::-:-::::::::-:::-:-:-:::-::-::::::::-:-:::-::::-:::::::-:::-:::::: :::::::-::::::::t::::: ::::::::::::::::::: :::::::::::::-:~:::-:::::::::::j-:-:::-::-::-::-::::- ·::::::-:-::::::: :-:-::::: .............. ~ . . 1850 1800 ....................................... , 1 .............. .. ........... , ..... .............. . ............. .. ......... - -. 0 10 20 30 40 50 60 70 80 Depth, cm FIGURE 9.23 Magnetic and chemical stratigraphy from a hummock peat core (1982) from Regent Street Bog, Fredricton, New Brunswick; dating by increments. IRM is isothermal remanent magnetism in units of Gauss, centimeters cubed per gram. (From F. Oldfield, University of Liverpool, U. K., and K. Tolonen, University of Oulu, Oulu, Finland, unpublished data.)

402 many physical, chemical, and biological processes make precise reconstruction of the deposition history of metals and other pollutants difficult. DIRECT COMPARISON OF DIATOM AND CHEMICAL DATA We have screened approximately 60 lakes in the United States for sediment data relating to pH reconstructions based on diatoms and about 40 lakes for sediment metals. Both chemical and biological data as well as lead-210 dating are available for 18 lakes, permitting direct comparison of diatom and chemical interpretations (Table 9.4). Tolonen and Jaakkola (1983) and Davis et al. (1983) also present comparisons of this type. Four of the lakes are at high elevation in the Rocky Mountain National Park in an area not currently receiving high levels of acid precipitation (Gibson et al. 1984). The remaining lakes in the sample are in the New England-New York region. Most of the lakes are low-alkalinity oligotrophic lakes and thus are not a representative sample of the lakes in the Northeast. In addition, the sample is biased toward lakes sensitive to atmospheric deposition of acids, and most were selected on the basis of being little disturbed and are thus at high altitude and remote. Diatom stratigraphy of the Rocky Mountain lakes shows no evidence of acidification (D. R. Beeson, U.S. Park Service, Fort Collins, Colorado, personal communication, 1984; Appendix Table E.1). The sediment chemistry of these lakes (Baron et al. 1984; Figure 9.24) indicates an influence of mining activities in the region as shown by TABLE 9.4 Lakes for Which Data on Diatoms, Sediment Chemistry, and Dating Are Available for Analysis Adirondack Park, N.Y. N. New England Rocky Mountain National Park, Colo. Big Moose Lake Deep Lake Panther Lake Sagamore Lake Woods Lake Branch Pond, Vt. E. Chairback Pond, Me. Klondike Pond, Me. Ledge Pond, Me. Mountain Pond, Me. Lake Solitude, N.H. Speck Pond, Me. Tumbledown Pond, Me. Unnamed Pond, Me. Emerald Lake Lake Haiyaha Lake Husted Lake Louise NOTE: For more detailed data on these lakes, see Appendix E, Tables E.1-E.5.

403 LAKE HUSTED ORGANIC CARBON A. 10 20' 4O O 10 20 30 40 Pb-21 0 TiO2 O- 0 _ 202- 1' 10 it: 30 1 30 I. 40~ 40 - _ ,, . , . . ., , _ , . . ~ · 0 150 0.0 0.6 pCi/g DRY PCT. IGN. CaO n MgO E it.| ,~ I, ~ ~ 'L I ., 1 ' 0.0 PCT. IGN. 1.0 0.0 1.5 PCT. IGN. MnO Pb Zn Cu O _ ._ O-, . _ 0~! 0- ..~ 20 10- OFF 20 '3 20 30 30 ~ 30~, . 30 If 40 40- i 40 1 ' 40~ 0.0 0.5 0 300 0 400 0 ! PCT. IGN. PPM IGN. PPM IGN. PPM IGN. A12O3 ° -L- -- 3s . , , , 1 0 ~=:~ 1 —- }l" .a, , I ~ 1 Her _ ~ 20~ 30 h¢~= A, I 404 of. ~ ~ I I . I .... 0 15 PCT. IGN FeO 0 14 '- - --- . 10 - 20 - 30~ 40~ ,.,.. 0 15 PCT. IGN 50 CONCENTRATION VERSUS DEPTH (cm) FIGURE 9.24 Sediment chemistry of Lake Husted, Rocky Mountain National park. The recent increase in lead is indicative of mining activities in the region. See caption, Figure 9.12, for definitions. (From Baron et al. 1984.) Diatom-inferred pH for this sediment core has been stable for at least the past 150 years, fluctuating around 6.6 to 6.8 pH units (see Appendix Table E.5; Beeson 1984). elevated lead concentrations but no history of fossil- fuel-related atmospheric inputs. This is indicated by the lack of elevated concetrations of certain diagnostic elements such as zinc. Surface water chemistry of lakes from this region indicates little or no acidification from excess atmospheric sulfate (see Appendix Table E.3; Wright 1983, Gibson et al. 1984). All study lakes in the northeastern United States have evidence of increased atmospheric deposition of metals, such as lead and copper, that are associated with combustion of fossil fuels and mining activities, but not

404 lyb 1 900 l~sSO ~ 18005 1 7so o 1950 - 1 900 - 1850 1 800 1 750 0.0 0.5 PCT. IGN. K2O 2000 - - 1950 - - 1 900 MgO 2000 1950 - 1850 - 1 800 - 17An. B=i! _. r -~r ~ - ~ L ~ ~ '` 1 1 1, ........ o · 900 1850 1 800 1 7~n - - - 1 i ~...... ~ ~v 0 30 PCT. IGN. A12O3 2000 - 1950 - 1 900 - 1850 1 800 1 750 I . _ 3_ _. _. 2000 - 1950 1 900 ' 1850 ' 1 800 1 750 "L · '~v · ,_,v I [JV 0.0 2.0 0 1 5 0 700 PCT. IGN. PCT. IGN. PPM IGN. CONCENTRATION VERSUS (YEAR) - 2000 MnO 1950 1 900 1850 1 800 1 750 - , ~ ;,v 0.0 0.1 0.0 0.5 PCT. IGN. PCT. IGN. ORGAN I C 2000 H2O 2000 CARBON 2000 Pb-210 2000 CaO _ ~1115 ]_ A_~ _ 1950_ 1950 ~ 1950 ~ ~ ~ _ _ 1 900 i_ 1 900 ~ 1 900 _ _ ~ ~ _ ~ 1850 _. 1850 - 1 1850 _ _ ~ ~ - ., ., ., 1800q ~ 1800- 1800 ~ `'. 1 750 1 750 1 750 100 0 50 0 50 0.0 1.5 PCT. WET PCT. DRY pCi/g DRY PCT. IGN. TiO2 Na2O Lr. ual_ 1 2000 1950~ 1 900 -_ 1 900 -_ _ ~ __ _ 1850 ~ I ~ 1~ l 1850_ 1800] tit 1800 L-::~ 's 1 . 1 1950~ 1 1 ~ 1 ...., .,, . 1 /bU -—- 0.0 1.0 PCT. IGN. 2000 - 1950- - 1900 1850 ~ 1800- 1 7~.^. _ Zn Cu 2000- 1950 1900 1850 ~ 1 800 ~ _ ., ~v 1 750 0 600 0 1 00 PPM IGN. PPM IGN. - l l ~ l l l l l l FIGURE 9.25 Concentrations versus year for various chemical species in Big Moose Lake (core 2). See caption, Figure 12, for definitions. Unpublished data from D. Charles (Indiana University) and S. Norton (University of Maine). all show evidence of recent acidification. Three case studies are presented here with comparative data. Big Moose Lake, New York, contains biological and chemical evidence that is indicative of acidification. Sediment from Speck Pond, Maine, has seemingly contradictory evidence. Panther Lake, New York, although receiving deposition of strong acids, shows little evidence of acidification. For Big Moose Lake, analyses of data on diatoms, chrysophytes, and sediment chemistry all indicate recent acidification (Appendix Table E.5; Figures 9.5 to 9.7, and 9.25). Diatom-inferred pH declines beginning around 1950, and falls rapidly in the 1960s. Changes in the

405 6~ - ~ 4 C: L1J a: By o ~ 3 o UJ 2 1 \ \~,,,\ 2000 1950 1900 1850 1800 1 750 YEAR FIGURE 9.26 Deposition rate of zinc versus year in Big Moose Lake (core 2). Unpublished data from S. Norton, University of Maine, Orono.

406 abundance of chrysophyte taxa (Figure 9.7) indicate the same pattern of pH change. The diatom-inferred pa also suggests that pH had been declining slowly since at least the early 1800s, possibly resulting from long-term gradual acidification from natural processes. Sediment chemistry data (Figures 9.25 and 9.26) are consistent with the diatom-inferred pH profiles, although changes do not generally occur at the same time. Both lead and zinc concentrations begin to increase in the sediment intervals dated late 1800s, as they do for most other lakes in the northeastern United States. The zinc-profile peak is associated with a lead-210 date of 1965-1970. The diatom- inferred pH for that time interval is 5.0. The zinc peak follows the onset of rapid pH decline by approximately 10 to 15 years. However, calculations of zinc deposition rate indicate a leveling of influx around 1950 (see Figure 9.26). This is consistent with the hypothesis that lower pH is associated with processes resulting in reduced influx of zinc to the sediment. Diagenesis may also contribute to the subsurface maximum in zinc concentration. Concentrations of calcium oxide (CaO) and manganese oxide (MnO) indicate long-term acidification consistent with an inferred early pH decrease. The titanium dioxide tTiO2), sodium oxide (Na2O) and potassium oxide (K2O) concentrations increase in the late 1800s and early 1900s, possibly in response to greater erosion of inorganic particles following the selective logging and the forest fire (1903) that occurred during this time. No significant pH shift is associated with these events. Speck Pond is an example of a lake with seemingly conflicting interpretations for microfossil evidence and chemical stratigraphy. The chemical stratigraphy of Speck Pond (S. A. Norton and R. B. Davis, University of Maine, unpublished data; lead and zinc profiles in Davis and Anderson 1984) is very similar to that of Jerseyfield Lake (Figure 9.12). The pH reconstruction for Speck Pond (Figure 9.8) suggests that the lake has been only slightly acidified if at all; the pH has remained in the range 5.2 to 4.8 for at least the past 100 years. Whereas the chemical profile of lead in Speck Pond is typical for nearly all lakes in the northeastern United States, the zinc profile has been interpreted by S. Norton (unpub- lished) to be caused by acidification of some parts of the ecosystem. Neither in situ leaching nor leaching of sedimenting particles (Kahl and Norton 1983, Schindler et al. 1980) by an acidifying water column is reflected in

407 the diatom data. Leaching by acid deposition of zinc and other metals bound to detritus before entry to the lake and subsequent sedimentation of detrital particulate matter is a plausible explanation for the impoverished concentration of sedimentous zinc. This explanation assumes that little reabsorption of dissolved zinc on sedimenting particles occurs in the lake. The present chemistry of Speck Pond is dominated by hydrogen ion, aluminum, calcium, and sulfate (Appendix Tables E.2 and E.3). No sources of sulfate have been identified in the watershed. Thus, before the input of excess sulfate, an anion must have been present to balance the hydrogen ions and some unknown amount of calcium and aluminum. Bicarbonate alkalinity would be virtually zero at pH = 5, but Lazerte and Dillon (1984) and others have postulated the importance of organic acids in determining lake acidity. One hypothesis is that organic acidity is replaced by sulfate-balanced acidity with a corresponding decline in dissolved organic carbon and increased water clarity (Davis et al. 1985), and an increased level of aluminum uncompleted by organic anions. This untested model would explain the observations of relatively invariant diatom-based pH with a concurrent decline of sediment deposition rates for calcium, manganese, and zinc (relative to TiO2 and A12O3) caused by leaching of particulates on the land before deposition; increased water clarity; and hydrogen-, calcium-, aluminum-, and sulfate-rich waters. Liming experiments in Sweden (H. Hultberg, Swedish Water and Air Pollution Research Institute, Gothenburg, personal com- munication) and in Norway (Wright 1984) have shown that when the pH of lakes is raised the reverse of some of these processes occurs (those involving diatoms, water clarity, dissolved organic carbon, and uncompleted aluminum). For Panther Pond, diatom-inferred pH (Appendix Table E.5; Del Prete and Galloway 1983) is nearly constant (around 6.5) throughout most of the core.* Concentrations *Analysis of diatom assemblages suggests a decline in pH of about 0.3 to 0.4 unit near the top of the sediment core based on the reduction in percent of euplanktonic diatoms within the core, a decrease in Cyclotella stelligera (indifferent), and an increase in Melosira distans var. distans (acidophil). However, reexamination

408 10- 20 - 30 - 40 - 1 80 100 PCT. WET Na2O 301 n:- 0 60 PCT. DRY ORGAN I C H2O CARBON Pb-210 O . O , 0W 0 10 20 30 . l ,- 30 ~ ._ , . 40 ~ - 0 50 0.0 0.5 pCi/g DRY PCT. IGN. K2O CaO MgO 20~ 234O~,, 20~ 234Oo~ 0.0 0.5 0.0 1.0 0.0 2.0 0.0 0.5 PCT. IGN. PCT. IGN. PCT. IGN. PCT. IGN. 0 10 20 An r Pb U 10 - 20 - 30 - 40 L .. .......... 0.0 0.8 0 500 PCT. IGN. PPM IGN. 0 1000 PPM IGN. CONCENTRATION VERSUS DEPTH (cm) TiO2 ~ A12O3 u 10 20 30 40 O- 10 - . . 20 - 30 404 O 10 - 20 301 404 0 50 PPM IGN. 1~........ 0 15 PCT. IGN. 0 10 PCT. IGN. FIGURE 9.27 Concentration versus depth for various chemical species in Panther Lake. See caption, Figure 9.12, for definitions. (From Norton 1984.) of lead and zinc (Figure 9.27) decrease toward the sediment surface. Based on these declines plus the shape of the lead-210 curve, it appears that the atmospheric inputs of lead-210, lead and zinc are diluted by increased deposition of major elements such as silicon dioxide. There is no decline of zinc, calcium, and manganese relative to other metals that are not mobilized by significantly lower pH. Deposition rates (not shown) of of the diatom assemblages at the top of the Panther Lake core, in light of recent taxonomic revisions, indicates a significant percent of the Melosirae should be reclassified as indifferent rather than acidophilic, and the diatom-inferred pH decrease may not be as great as first determined (D. Charles, unpublished data).

409 such metals as titanium, manganese, aluminum, sodium, and potassium increase dramatically from the late 1800s to the early 1900s, when logging occurred in the watershed. Because Panther Lake currently has an alkalinity about 150 peq/L (it was probably greater earlier), it is relatively well buffered against changes that might result from changing atmospheric deposition or watershed disturbance. The sediment chemistry data are consistent with the relatively constant diatom-inferred pH. Although the chemistry data indicate an influx of material associated with fossil fuel combustion and watershed disturbance, there is no strong evidence for increased lake-water acidification. CONCLUSIONS Diatoms and Chrysophytes 1. Diatom and chrysophyte data for 31 lakes were evaluated to assess regional trends in lake acidification. The lakes in each region (Adirondack Mountains, eleven; New England, ten; eastern Canada, six; and the Rocky Mountains, four) were generally not representative of the range of chemistries displayed by lakes in those regions. 2. There is evidence that certain poorly buffered lakes have become more acidic recently. The lakes for which evidence of acidification is strongest are in the Adirondack Mountains, New York. Of the eleven lakes for which there are data, six of the seven lakes with a current pH at or below 5.0 show evidence of recent acidification, while one is a naturally acidic bog lake. None of the four lakes with current pH above about 5.0 shows any obvious evidence of substantial pH decline. Where dating is available, evidence indicates that the most rapid pH changes (decrease of 0.5 to 1.0 pH unit) occurred between 1930 and 1970. For some lakes there have been watershed disturbances that make sediment data more difficult to evaluate. 3. The diatom data for New England lakes indicate either a slight or no decrease in pH. Diatom data for lakes in eastern Canada indicate no change in pH (four lakes) or a significant decrease, which is probably caused primarily by local smelting activities (two lakes). There is no evidence to indicate declines in pH in the Rocky Mountain lakes.

4 ~ o 4. Before 1800, several lakes in the Adirondack Mountains and New England had pH values less than 5.5. These lakes now have a pH of only 0.1 to 0.4 pH unit lower and total aluminum concentrations greater than 100 ug/L. Because of the potential importance of organic acid and aluminum buffering, a small decline in pH could be associated with significant decreases in alkalinity. 5. Analysis of the Adirondack lake/watershed data sets indicates that the recent rapid declines in lake-water pH are most likely related to increased acid deposition. However, watershed disturbances may also play a role. 6. Analysis of the three lakes in the Adirondacks (Big Moose Lake, Deep Lake, and Upper Wallface Pond) for which data are adequate for the timing of acid-deposition- related events to be assessed suggests that the onset of decline began in the 1930s to 1950s. 7. Analysis of sediment diatom and chrysophyte assemblages is the best paleolimnological technique currently available for reconstructing past lake-water pH. However, there are important uncertainties in the technique, and further research is necessary to improve procedures and interpretation of the results. Geochemistry 1. Lead deposition from the atmosphere started increasing sometime in the middle to late 1800s, as shown by increasing concentrations in sediments and increasing sediment-deposition rates of lead, independent of most other elements. 2. The history of zinc sedimentation roughly parallels that of lead. However, in recent sediment in acidic (pH < 5.5) lakes, concentrations and apparent deposition rates of zinc and other easily mobilized metals decrease toward the surface of the sediment. 3. Copper and cadmium increase less than lead or zinc. Other metals commonly not measured, such as chromium, arsenic, and antimony, also increase in this same period. 4. The concentration and the deposition rate of vanadium increase around 1940. 5. PAHs and variously shaped carbonaceous particules related to fossil fuel consumption are found in lake sediments. Their concentrations and deposition rates are interpretable in terms of estimated emissions resulting from various processes.

411 6. Trace metals, PAHs, and magnetic spherules are present in ombrotrophic peat profiles. Their concen- trations are elevated in peat dating from current times to more than 100 years old. Trends in deposition can be estimated only for cores from hollows in the bogs and then only crudely. The abundance of spherules appears to mimic most closely the known emission histories in the local (<100-km) region. Interpretations of lake acidity and atmospheric deposition changes based on stratigraphic analysis of both diatom assemblages and sediment chemistry are generally consistent. REFERENCES Alhonen, P. 1967. Paleolimnological investigations of three inland lakes in southwestern Finland. Acta Bot. Fenn. 76:1-59. Alhonen, P. 1968. On the late-glacial and early post- glacial diatom succession in Loch of Park, Aberdeenshire, Scotland. Mem. Soc. Fauna Flora Fenn. 44:13-20. Almer, B., W. Dickson, C. Ekstrom, E. Hornstrom, and U. Miller. 1974. Effects of acidification on Swedish lakes. Ambio 3:30-36. Almer, B., W. Dickson, C. Ekstrom, and E. Hornstrom. 1978. Sulfur pollution and the aquatic ecosystem. Pp. 271-311 of Sulfur in the Environment, Part II, Ecological Impacts, J. O. Nriagu, ed. New York: John Wiley and Sons. Andersen, S. T. 1966. Interglacial vegetational succession and lake development in Denmark. Paleobotanist 15:117-127. Anderson, D. S., R. B. Davis, and F. Berge. In press. Relationships between diatom assemblages in lake surface-sediments and limnological characteristics in southern Norway. Hydrobiologia. Andersson, G., S. Fleischer, and W. Graneli. 1978. Influence of acidification on decomposition processes in lake sediment. Verh. Int. Ver. Limnol. 20:802-807. Appleby, P. G., and F. Oldfield. 1978. The calculation of lead-210 dates assuming a constant rate of supply of unsupported 210Pb to the sediment. Catena 5:1-8. Arzet, K., D. Krause-Dellin, and C. Steinberg. 1985. Acidification of selected lakes in the Federal

412 Republic of Germany as reflected by subfossil diatoms, Cladoceran remains, and sediment chemistry. In Diatoms and Lake Acidity, J. P. Smol, R. W. Battarbee, R. B. Davis, and J. Merilainen, eds. The Hague, The Netherlands: W. Junk. Baron, J. 1983. Comparative water chemistry of four lakes in Rocky Mountain National Park. Water Res. Bull. 19:897-902. Baron, J., D. R. Beeson, S. A. Zary, P. M. Walthall, W. L. Lindsay, and D. M. Swift. 1984. Long-term research into the effects of acidic deposition in Rocky Mountain National Park: Summary Report, 1980-1984. U.S. Park Service. Tecn. Rep. No. 84-~0~0-2. Battarbee, R. W. 1979. Diatoms in lake sediments. Pp. 177-225 of Paieohydrological Changes in the Temperate Zone in the Last 15000 Years, Subproject B. Lake and Mire Environments, B. E. Berglund, ed. Vol. II, Lund University, Sweden. Battarbee, R. W. 1984. Diatom analysis and the acidification of lakes. Phil. Trans. R. Soc. London 305:451-477. Batterbee, R. W., and R. J. Flower. 1984. The inwash of catchment diatoms as a source of error in the sediment-based reconstruction of pH in an acid lake. Limnol. Oceanogr. 29:1325-1329. Baxter, M. S., J. G. Farmer, I. G. McKinley, D. S. Swan, and W. Jack. 1981. Evidence of the unsuitability of gravity coring for collecting sediment in pollution and sedimentation rate studies: Environ. Sci. Technol. 15:843-846. Beamish, R. J., and H. H. Harvey. 1972. Acidification of the La Cloche Mountain lakes, Ontario, and resulting fish mortalities. J. Fish. Res. Board Can. 29:1131-1143. Beeson, D. R. 1984. Historical fluctuations in lake pH as calculated by diatom remains in sediments from Rocky Mountain National Park. Bull. Ecol. Soc. Am. 65:135. Berg, C. O. 1966. Middle Atlantic States. Pp. 191-237 of Limnology in North America, D. G. Frey, ed. Madison, Wisc.: University of Wisconsin Press. Berge, F. 1976. Kiselalger og pH i noen elver og innsj~er i Agder og Telemark. En sammenliKning me~lom arene 1949 og 1975. [Diatoms and pH in some rivers and lakes in Agder and Telemark (Norway): a comparison between the years 1949 and 1975.] Sur Nedb~rs Virkning Pa Skog Og Fisk. IR18/76. Norwegian Forest Researcn Institute, Aas, Norway.

413 Berge, F. 1979. Kiselalger Agder og Hordaland. Sur Fisk. r Di og pH i noen innsjoer i og noroa~ana. Our Nedb~rs Virkning Pa Skog Og [Diatoms and pH in some lakes in the Agder and Hordaland Counties, Norway.] IR42/79. Norwegian Forest Research Institute, Aas, Norway. Berge, F. 1983. Diatoms as indicators of temporal pH trends in some lakes and rivers in southern Norway. Nova Hedwigia 73:249-265. Berglund, B., and N. Maimer. 1971. Soil conditions and late-glacial stratigraphy. Geol. Foeren. Stockholm Foerh. 93:575-586. Bradstreet, T. E., and R. B. Davis. 1975. Mid-postglacial environments in New England with emphasis on Maine. Arct. Anthropol. 12:7-22. Brakke, D. F. 1984. Acidification history and prediction using biological indices. Verh. Int. Ver. Limnol. 22:1391-1395. Brugam, R. B. 1980. Postglacial diatom stratigraphy of Kirchner Marsh, Minnesota. Quat. Res. (N.Y.) 13:133-146. Brugam, R. B. 1983. The relationship between fossil diatom assemblages and limnological conditions. Hydrobiologia 98:223-235. Brugam, R. B. 1984. Holocene paleolimnology. Pp. 208-221 of Late Quaternary Environments of the United States, Vol. 2: The Holocene, H. E. Wright, Jr., ed. Minneapolis, Minn.: University of Minnesota Press. Buckley, D. E., and R. E. Cranston. 1971. Atomic absorption analyses of 18 elements from a single decomposition of alumino-silicate. Chem. Geol. 7:273-284. Campbell, P. G. C., and A. Tessier. 1984. Paleo- limnological approaches to the study of acid deposition: metal partitioning in lacustrine sediments. Environmental Protection Agency Final Report on the Workshop, Progress and Problems in tne Paleolimnological Analysis of the Impact of AcidiC Precipitation and Related Pollutants on Lake Ecosystems, S. A. Norton, ed. 40 pp. Carignan, R., and J. O. Nriagu. 1985. Trace metal deposition and mobility in the sediments of two lakes near Sudbury, Ontario. Geochim. Cosmochim. Acta 49:1753-1764. Carignan, R., and A. Tessier. 1985. Zinc deposition in acid lakes: the role of diffusion. Science 228:1524-1526.

414 Charles, D. F. 1982. Studies of Adirondack Mountain (N.Y.) lakes: limnological characteristics and sediment diatom-water chemistry relationships. Doctoral dissertation, Indiana University, Bloomington, Ind. Charles, D. F. 1984. Recent pH history of Big Moose Lake (Adirondack Mountains, New York, USA) inferred from sediment diatom assemblages. Int. Ver. Theor. Angew. Limnol. Verh. 22:559-566. Charles, D. F. 1985a. Relationships between surface sediment diatom assemblages and lakewater characteristics in Adirondack lakes. Ecology 66:994-1011. Charles, D. F. 1985b. A new diatom species, Fragiiaria acidobiontica, from acidic lakes in northeastern North America. In Diatoms and Lake Acidity, J. P. Smol, R. W. Battarbee, R. B. Davis, and J. Merilainen, eds. The Hague, The Netherlands: W. Junk. CAolnoky, B. J. 1968. Die Okologie der Diatomeen in Binnengewasser. J. Federal Republic of Germany: Lehre, Cramer. Cleve, P. T. 1891. The diatoms of Finland. Acta Soc. Fauna Flora Fenn. 8:1-68. Clymo, R. S. 1963. Ion exchange in Sphagnum and its relation to bog ecology. Ann. Bot. 27:309-324. Clymo, R. S. 1967. Experiments on breakdown of Sphagnum in two bogs: J. Ecol. 53:747-757. Clymo, R. S. 1978. A model of peat bog growth: Pp. 187-223 of Ecological Studies, O. W. Heal and D. F. Perkins, eds. Vol. 27. Berlin: Springer-Verlag. Colingsworth, R. F., M. H. Hohn, and G. B. Collins. 1967. Post-glacial physiochemical conditions of Vestaburg Bog, Montcalm County, Michigan, based on diatom analysis. Pap. Mich. Acad. Sci. Arts Lett. 52:19-30. Colquhoun, J., W. Krester, and M. Pfeiffer. 1984. Acidity status update of lakes and streams in New York State. New York State Department of Environmental Conservation, Albany, N.Y. 139 pp. Conger, P. S. 1939 . The contribution of diatoms to the sediments of Crystal Lake, Vilas County, Wisconsin. Am. J. Sci. 237: 324-340 . Crabtree, K. 1969. Post-glacial diatom zonation of limnic deposits in north Wales. Mitt. Int. Ver. Theor. Angew Limnol. 17: 165-171. Crecelius, E. A. 1975. The geochemical cycle of arsenic in Lake Washington and its relation to other elements. Limnol. Oceanogr. 20: 441-451. .

415 Crocker, R. L., and J. Major. 1955. Soil development in relation to vegetation and surface age at Glacier Bay, Alaska. J. Ecol. 43:427-448. Damman, A. W. H. 1978. Distribution and movement of elements in ombrotrophic peat bogs. Oikos 30:480-495. Davis, R. B., and D. S. Anderson. 1984. Assessment of trends based on paleolimnological technique. Pp. 4-99 -- 4-105 of The Acidic Deposition Phenomenon and Its Effects: Critical Assessment Review Papers, Vol. 2, A. P. Altshuller and R. A. Linthurst, eds. Washington, D.C.: U.S. Environmental Protection Agency. Davis, R. B., and D. S. Anderson. 1985. Methods of pit calibration of sedimentary diatom remains for reconstructing history of pH in lakes. Hydrobiologia 120:69-87. Davis, R. B., and F. Berge. 1980. Atmospheric deposition in Norway during the last 300 years as recorded in SNSF lake sediments II. Diatom stratigraphy and inferred pH. Pp. 270-271 of Ecological Impact of Acid Precipitation. Proceedings of an international conference, Sandefiord, Norway. D. Disables and A. Tollan, eds. Sur Nedb~eSrs Virkning Pa Skog Og Fisk (SNSF) Project, Oslo. Davis, R. B., and S. A. Norton. 1978. Paleolimnologic studies of human impact on lakes in the United States, with emphasis on recent research in New England. Poll Arch. Hydrobiol. 25:99-115. Davis, R. B., T. E. Bradstreet, R. Stuckenrath, Jr., and H. W. Borns. 1975. Vegetation and associated environments during the past 14,000 years near Moulton Davis, R. B., S. A. Norton, C. T. Hess, and D. F. Brakke. Davis, R. B., S. A. Norton, C. T. Hess, and D. F. Brakke. 1983. Paleolimnological reconstruction of the effects of atmospheric deposition of acids and heavy metals on the chemistry and biology of lakes in New England and Norway. Hydrobiologia 103:113-123. Davis, R. B., C. T. Hess, S. A. Norton, D. W. Hanson, K. D. Hoagland, and D. S. Anderson. 1984. 137cs and 210Pb dating of sediments from soft water lakes in New England (U.S.A.) and Scandinavia, a failure of 137Cs dating. Chem. Geol. 44:151-185. Davis, R. B., D. S. Anderson, and F. Berge. 1985. Loss of organic matter, a fundamental process in lake acidification: paleolimnological evidence. Nature 316:436-438.

416 Davis, R. B., D. S. Anderson, D. F. Charles, and J. N. Galloway. Two-hundred year pH history of Woods, Sagamore, and Panther Lakes in the Adirondack Mountains, New York State, USA. (In preparation.) Dearing, J. A. 1983. Changing patterns of sediment accumulation in a small lake in Scania, southern Sweden: Hydrobiologia 103:59-64. Del Prete, A. 1912. Postglacial diatom changes in Lake George, New York. Ph.D. dissertation, Rensalear Polytechnic Institute, Troy, N.Y., 110 pp. Del Prete, A., and J. N. Galloway. 1983. Temporal trends in the pH of Woods, Sagamore and Panther lakes as determined by an analysis of diatom populations. In The Integrated Lake-Watershed Acidification Study: Proceedings of the ILWAS Annual Review Conference, Palo Alto, Calif.: Electric Power Research Institute. Del Prete, A., and C. Schofield. 1981. The utility of diatom analysis of lake sediments for evaluating acid precipitation effects on dilute lakes. Arch. Hydrobiol. 91:332-340. Delorme, L. D., S. R. Esterby, and H. Dutnie. 1984. Prehistoric pH trends in Kejimku~ik Lake, Nova Scotia. Int. Rev. Gesam. Hydrobiol. 69:41-55. Delorme, L. D., H. C. Duthie, S. R. Esterby, S. M. Smith, and N. S. Harper. In preparation. Prenistoric inferred pH changes in Batchawana Lake, Ontario from sedimentary diatom assemblages. Dickman, M., and J. Fortescue. 1984. Rates of lake acidification inferred from sediment diatoms for 8 lakes located north of Lake Superior, Canada. Int. Ver. Theor. Angew. Limnol. Verh. 22:1345-1356. Dickman, M., S. Dixit, J. Fortescue, R. Barlow, and J. Terasmae. 1983. Diatoms as indicators of the rate of lake acidification. Water Air Soil Pollut. 21:227-241. Dickson, W. 1980. Properties of acidified waters. Pp. 75-83 of The Ecological Impact of Acid Precipitation. Proceedings of an international conference, Sandefjord, Norway. D. Drabl~s and A. Tollan, eds. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF) project, Oslo. Digerfeldt, G. 1972. The post-glacial development of Lake Trummen. Regional vegetation history, water level changes, and paleolimnology. Folia Limnol. Scand. 16:1-104. Digerfeldt, G. 1975. The post-glacial development of Ranviken Bay in Lake Immein. III. Paleolimnology. Geol. Foeren. Stockholm Foerh. 97:13-28.

417 Digerfeldt, G. 1977. The Flandrian development of Lake Flarken. Regional vegetation history and paleolimnology. Department of Quaternary Geology, University of Lund, Sweden, Report 13. 101 pp. Dillon, P. J., and R. D. Evans. 1982. Whole-lake lead burdens in sediments of lakes in southern Ontario, Canada. Hydrobiologia 91:121-130. Driscoll, C. T. 1980. Chemical characterization of some dilute acidified lakes and streams in the Adirondack region of New York State. Ph.D. dissertation, Cornell University, Ithaca, N.Y. Driscoll, C. T., and R. M. Newton. 1985. Chemical characteristics of "acid sensitive" lakes in the Adirondack region of New York. Environ. Sci. Tecnnol. 19:1018-1024. Driscoll, C. T., J. P. Baker, J. J. Bisogni, and C. L. Schofield. 1984. Aluminum speciation and equilibria in dilute acidic surface waters of the Adirondack region of New York State. Pp. 55-75 of Geological Aspects of Acid Deposition, O. Bricker, ed. Boston, Mass.: Butterworth Publishers. Edginton, D. N., and J. A. Robbins. 1976. Records of lead deposition in Lake Michigan sediments since 1800: Environ. Sci. Technol. 10:266-274. Esterby, S. R., and A. H. El-Shaarawi. 1981a. Likelihood inference about the point of change in a regression regime. J. Hydrol. 53:17-30. Esterby, S. R., and A. H. El-Shaaraw i. 1981b. Inference about the point of change in a regression model. J. Appl. Stat. 30:277-285. Evans, G. H. 1970. Pollen and diatom analysis of late-quaternary deposits in the Blelham Basin, North Lancashire. New Phytol. 69:821-874. Evans, G. H., and R. Walker. 1979. The Late-quaternary history of the diatom flora of Llyn Clyd and Llyn Glas, two small oligotrophic high mountain tarns in Snowdonia (Wales). New Phytol. 78:221-236. Evans, H. E., P. J. Smith, and P. J. Dillon. 1983. Anthropogenic zinc and caamium burdens in sediments of selected southern Ontario la~es. Can. J. Fish. Aquat. Sci. 40:570-579. Evans, R. D., and P. J. Dillon. 1982. Historical changes in anthropogenic lead fallout in southern Ontario. Hydrobiologia 91:131-137. Flower, R. J., and R. W. Battarbee. 1983. Diatom evidence for recent acidification of two Scottish lochs. Nature 305:130-133.

418 Flower, R. J., and R. W. Battarbee. 1985. The morphology and biostratigraphy of Tabellaria quadriseptata (Bacillariophycene) in acid waters and lake sediments in Galloway, Southwest Scotland. Brit. Phycol. J. 20:69-78. Foged, N. 1969. Diatoms in a post-glacial core from the bottom of the Lake Grane Langs0`, Denmark. Dan. Geol. Foren. 19:237-256. Foged, N. 1972. The diatoms in four postglacial deposits in Greenland. Medd. Groen. 194:1-67. Ford, M. S. (J.) 1984. The influence of Ethology on ecosystem development in New England: a comparative paleoecological study. Doctoral dissertation, University of Minnesota, Minneapolis, Minn. Ford, M. S. (J.) 1985. The recent history of a naturally acidic lake (Cone P., N.H.). In Diatoms and Lake Acidity, J. P. Smol, R. W. Battarbee, R. Davis, and J Merilainen, eds. The Hague, The Netherlands: W. Junk. (In press.) Fortescue, J. A. C. 1984. Interdisciplinary research for an environmental component (acid rain) in regional geochemical surveys (Wawa area). Algoma District, Ontario Geological Survey Map 80 713, Geochemical Series, compiled 1983. Fortescue, J. A. CO' I. Thompson, M. Dic~man, and J. Terasmae. 1981. Multidisciplinary followup of regional pH patterns in lakes north of Lake Superior, District of Algoma, Ontario Geological Survey OPR 5342. Part I, 134 pp., 6 tables, 65 figures. Fortescue, J. A. C., I. Thompson, M. Dickman, and J. Terasmae. 1984. Multidisciplinary followup of regional pH patterns in lakes north of Lake Superior. District of Algoma, Ontario, Geological Survey OFR 5342. Part II, Information and data listings for lakes studied. 366 pp. Frey, D. G. 1969. The rationale of paleolimnology. Mitt. Int. Ver. Theor. Angew. Limnol. 17:7-18. Fritz, S. C., and R. E. Carlson. 1982. Stratigraphic diatom and chemical evidence for acid strip-mine lake recovery. Water Air Soil Pollut. 17:151-163. Galloway, J. N., and G. E. Likens. 1979. Atmospheric enhancement of metal deposition in Adirondack Lake sediments: Limnol. Oceanogr. 24:427-433. Galloway, J. N., J. D. Thornton, S. A. Norton, H. L. Volchok, and R. A. N. McLean. 1982. Trace metals in atmospheric deposition: a review and assessment: Atmos. Environ. 16:1677-1700.

419 Galloway, J. N., C. L. Schofield, N. E. Peters, G. R. Hendrey, and E. R. Altwicker. 1983. Effect of atmospheric sulfur on the composition of three Adirondack lakes. Can. J. Fish. Aguat. Sci. 40:799-806. Gasse, F., and F. Tekaia. 1983. Transfer functions for estimating paleoecological conditions (pH) from East African diatoms. Hydrobiologia 103:85-90. Gasse, F., J. F. Talling, and P. Kilham. 1983. Diatom assemblages in East Africa: classification, distribution and ecology. Rev. Hydrobiol. Tropic. 16:3-34. Gensemer, R. W., and S. S. Kilham. 1984. Growth rates of five freshwater algae in well-buffered acidic media. Can. J. Fish. Aquat. Sci. 41:1240-1243. Gibson, J. H., J. N e Galloway, C. Schofield, W. McFee, R. Johnson, S. McCorley, N. Dise, and D. Herzog. 1984. Rocky Mountain Acidification Study. U.S. Fish and Wildlife Service, Division of Biological Services, Eastern Energy and Land Use Team. FWS/OBS-80/40.17. 137 pp. Gorham, E. 1955. On the acidity and salinity of rain: Geochim. Cosmochim. Acta 7:231-239. Gorham, E. 1958. Atmospheric pollution by hydrochloric acid, Quart. J. R. Meteorol. Soc. 84:274-276. Gorham, E., P. M. Vitousek, and W. A. Reiners. 1979. The regulation of chemical budgets over the course of terrestrial ecosystem succession. Ann. Rev. Ecol. Syst. 10:53-84. Gorham, E., S. J. Eisenreich, J. Ford, and M. V. Santelmann. 1985. The chemistry of bog waters. Pp. 339-363 of Chemical Processes in Lakes, W. Stumm, ed. New York: Wiley-Interscience. Griffin, J. J., and E. D. Goldberg. 1981. Sphericity as a characteristic of solids from fossil fuel burning in a Lake Michigan sediment. Geochim. Cosmochim. Acta 45:763-769. Groet, S. S. 1976. Regional and local variations in heavy metal concentrations of bryophytes in the northeastern United States. Oikos 27:445-456. Hakanson, L. 1977. The influence of wind, fetch, and water depth on the distribution of sediments in Lake Vanern, Sweden. Can. J. Earth Sci. 14:397-412. Hanson, D. W., S. A. Norton, and J. S. Williams. 1982. Modern and paleolimnological evidence for accelerated leaching and metal accumulation in soils in New England caused by atmospheric deposition. Water Air Soil Pollut. 18:227-239.

420 Hase, A., and R. A. Hites. 1976. On the origin of polycyclic aromatic hydrocarbons in recent sediments: biosynthesis by anaerobic bacteria. Geochim. Cosmochim. Acta 40:1141-1143. Haworth, E. Y. 1969. The diatoms of a sediment core from Thea Tarn, Langdale. J. Ecol. 57:429-439. Haworth, E. Y. 1972. Diatom succession in a core from Pickerel Lake, northeastern South Dakota. Geol. Soc. Am. Bull. 83:157-172. Heit, M., Y. Tan, C. Klusek, and J. C. Burke. 1981. Anthropogenic trace elements and polycyclic aromatic hydrocarbon levels in sediment cores from two lakes in the Adirondack acid lake region. Water Air Soil Pollut. 15:441-464. HemOnd, H. F. 1980. Biogeochemistry of Thoreau's Bog, Concord, Massachusetts. Ecol. Monogr. 50:507-526. Holdren, G. R., Jr., T. M. Brunelle, G. Matisoff, and M. Wahlen. 1984. Timing the increase in atmospheric sulphur deposition in the Adirondack Mountains. Nature 311:245-248. Hongve, D. 1978. Buffering of acid lakes by sediments: Verh. Int. Ver. Limnol. 20:743-748. Hongve, D., and A. Erlandsen. 1979. Shortening of surface sediment cores during sampling. Hydrobiologia 65:283-287. Howells, G. D. 1983. Acid waters--the effect of low pH and acid associated factors on fisheries. Appl. Biol. 9:143-255. Hultberg, H., and I. Andersson. 1982. Liming of acidified lakes and streams--aspects on induced physical-chemical and biological changes. Water Air Soil Pollut. 18:311-331. Hustedt, F. 1927-1966. Die Kieselalgen. Rabenhorsts Kryptogamen-flora von Deutschland, Osterreich und der Schweiz. Band 7. Leipzig, Germany: Geest und Portig. Hustedt, F. 1939. Systematische und okologische Untersuchungen uber die Diatomeen-Flora von Java, Bali, und Sumatra nach dem Material der Deutschen Limnologischen Sunda-Expedition III. Die okologischen Factorin und ihr Einfluss auf die Diatomeenflora. Arch. Hydrobiol. Suppl. 16:274-394. Huttunen, P., and J. Merilainen. 1983. Interpretation of lake quality from contemporary diatom assemblages. Hydrobiologia 103:91-97. Iskandar, I. K., and D. R. Keeney. 1974. Concentration of heavy metals in sediment cores from selected Wisconsin lakes. Environ. Sci. Technol. 8:165-170.

421 Jacobson, G. L., Jr., and H. J. B. Birks. 1980. Soil development on recent end moraines of the Klutlan Glacier, Yukon Territory, Canada. Quat. Res. 14:87-100 Jeffries, D. S., R. G. Semkin, R. Neureuther, and M. D. Jones. 1983. Data Report: Major ion composition of lakes in the Turkey Lakes Watershed (January 1980-May 1982). Turkey Lakes Watershed, Unpublished Report No. TLW-83-11. 9 pp. Johannessen, M. 1980. Aluminum--a buffer in acidic waters. Pp. 222-223 of Ecological Impacts of Acid Precipitation, Proceedings of the International Conference, Sandefjord, Norway. D. Disables and A. Tolonan, eds. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF) (Acid Precipitation--Effects on Forest and Fish) project, Oslo. Johnson, D. W., and D. W. Cole. 1977. Sulfate mobility in an outwash soil in western Washington. Water Air Soil Pollut. 7:489-495. Johnson, N. M., C. T. Driscoll, J. S. Eaton, G. E. Likens, and W. H. McDowell. 1981. "Acid rain," dissolved aluminum and chemical weathering at tne Hubbard Brook Experimental Forest, New Hampshire. Geochim. Cosmochim. Acta 45:1421-1437. Johnston, S. E. 1980. A comparison of dating methods in laminated lake sediments in Maine. M.S. thesis, University of Maine, Orono, Me. 79 pp. Johnston, S. E., S. A. Norton, C. T. Hess, R. B. Davis, and R. S. Anderson. 1982. Chronology of atmospheric deposition of acids and metals in New England based on the record in lake sediments. Pp. 117-127 of Energy and Environmental Chemistry, L. H. Keith, ed. Ann Arbor, Mich.: Ann Arbor Science. J¢rgensen, E. G. 1948. Diatom communities in some Danish lakes and ponds. K. Dan. Vidensk. Selsk. Biol. Skr. 5:1-140. Junge, G. E., and R. J. Werby. 1958. The concentration of chloride, sodium, potassium, calcium and sulfate in rainwater over the United States. J. Meteorol. 15:417-425. Kahl, J. S., and S. A. Norton. 1983. Metal input and mobilization in two acid-stressed lake watersheds in Maine: Completion Report, Project A-053-ME. Land and Water Resources Center, University of Maine, Orono, Me. 70 pp. Kah1, J. S., S. A. Norton, and J. S. Williams. 1984. Chronology, magnitude and paleolimnological record of changing metal fluxes related to atmospheric .

422 deposition of acids and metals in New England. Pp. 23-35 of Geologic Aspects of Acid Deposition, O. P. Bricker, ed. Ann Arbor, Mich.: Ann Arbor Science. Kolbe, R. W. 1932. Grundlinien einer allgemeinen ~kologie der Diatomeen. Ergeb. Biol. 8:221-348. Larsen, R. J. 1984. Worldwide deposition of 90Sr through 1982. EML-430. Environmental Measurements Laboratory, Department of Energy. 168 pp. Lazerte, B., and P. Dillon. 1984. Relative importance of anthropogenic versus natural sources of acidity in lakes and streams of central Ontario. Can. J. Fish. Aquat. Sci. 41:1664-1677. Lazrus, A. L., E. Lorange, and I. P. Lodge. 1970. Lead and other ions in U.S. precipitation. Environ. Sci. Technol. 4:55-58. Lillie, R. A., and J. W. Mason. 1983. Limnological characteristics of Wisconsin lakes. Tech. Bull. No. 138. Wisconsin Department of Natural Resources. 116 pp. Livett, E. A., J. A. Lee, and J. H. Tallis. 1979. Lead, zinc, and copper analyses of British blanket peats. J. Ecol. 67:865-891. Livingstone, D. A., K. Bryan, Jr., and R. G. Leaky. 1958. Effects of an arctic environment on the origin and development of freshwater lakes. Limnol. Oceanogr. 3:192-214. Longmore, M. E., B. M. O'Leary, and C. W. Rose. 1981. Cesium-137 profiles in a near meromictic lake on Great Sandy Island (Fraser Island), Queensland, Australia. Paper presented at the Third International Symposium on Paleolimnology, Joensur, Finland. Lowe, R. L. 1974. Environmental requirements and pollution tolerance of freshwater diatoms. EPA-670/4-74-005. U.S. Environmental Protection Agency. Lundquist, G. 1924. Utvecklingshistoriska insjostudier i Syd-sverige. Sver. Geol. Unders. Arsb. Ser. C 330:1-129. Madsen, P. P. 1981. Peat bog records of atmospheric mercury deposition. Nature 293:127-130. Martin, C. W., D. S. Noel, and C. A. Federer. 1984. Effects of forest clearcutting in New England on stream chemistry. J. Environ. Qual. 13:204-210. Mayer, L. M., F. P. Liotta, and S. A. Norton. 1982. Hypolimnetic redox and phosphorus cycling in hypereutrophic Lake Sebasticook, Maine. Water Res 16:1189-1196. .

423 Merilainen, J. 1967. The diatom flora and the hydrogen- ion concentration of the water. Ann. Bot. Fenn. 4:51-58. Mitchell, M. J., D. H. Landers, and D. F. Brodowski. 1981. Sulfur constituents of sediments and their relationship to lake acidification. Water Air Soil Pollut. 16:351-359. Mitchell, M. J., M. B. David, and A. J. Uutala. 1985. Sulfur distribution in lake sediment profiles as an index of historical depositional patterns. - Hydrobiologia 121:121-127. National Atmospheric Deposition Program. 1984. National Acid Deposition Program Data Report, Fourth Quarter, 1982. Vol. 4, 227 pp. Neiboer, E., H. M. Ahmed, K. J. Puckett, and D. H. S. Richardson. 1972. Heavy metal content of lichens in relation to distance from a nickel smelter in Sudbury, Ontario. Lichenologist 5:292-304. New York State Department of Environmental Conservation. 1978. Lake and pond survey data sheets. Albany, N.Y. Nilsson, S. I., H. G. Miller, and J. D. Miller. 1982. Forest growth as a possible cause of soil and water acidification: an examination of the concepts. Oikos 39:40-49. Norton, S. A. 1983. Lake sediment studies: An historical perspective of the chronology and impact of atmospheric deposition. Report of the NATO/CCMS Pilot Study on Air Pollution Control Strategies and Impact Modeling. Norton, S. A. 1983. Sediment chemistry of four lakes in Rocky Mountain National Park, Colorado. Draft final report. Water Resources Field Support Laboratory. U.S. Department of the Interior, National Park Service, Colorado State University, Fort Collins, Colo. Norton, S. A. 1984. Paleolimnological assessment of acidic deposition impacts on select acidic lakes: EPA/NCSU Acid Precipitation Program. Final report to U.S. Environmental Protection Agency; contract no. APP-0119-1982, 127 pp. Norton, S. A. 1985. The sedimentary record of atmospheric pollution in Jerseyfield Lake, Adirondack Mountains, New York. In Acid Deposition--Environmental, Economic, and Policy Issues, D. D. Adams and W. Page, eds. New York: Plenum. Norton, S. A. Geochemistry of selected Maine peat bogs. Maine Geol. Surv. Bull. 34. 38 pp. (In press.)

424 Norton, S. A., and C. T. Hess. 1980. Atmospheric deposition in Norway during the last 300 years as recorded in S.N.S.F. lake sediment. 1. Sediment dating and chemical stratigraphy. Pp. 268-269 of Proceedings of International Conference on the Ecological Impact of Acid Precipitation. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF), Sandefjord, Norway. Norton, S. A., and R. F. Wright. 1984. Buffering by sediments during liming and reacidification of two acidified lakes in southern Norway. Paper presented at the Third International Symposium on Interaction between Sediment and Water, Geneva, Switzerland, Aug. 28-31. Norton, S. A., R. F. Dubiel, D. R. Sasseville, and R. B. Davis. 19 7 8 . Paleolimnologic evidence for increased zinc loading in lakes of New England, U.S.A. Verh. Int. Ver. Limnol. 20:538-545. Norton, S. A., C. T. Hess, and R. B. Davis. 1980. Rates of Accumulation of Heavy Metals in Pre- and Post-European Sediments in New England Lakes. Ann Arbor, Mich.: Ann Arbor Science. Norton, S. A., R. B. Davis, and D. F. Brakke. 1981. Responses of northern New England lakes to atmospheric inputs of acids and heavy metals. Completion Report Project A-048-ME. U.S. Department of the Interior. Land and Water Resources Center, University of Maine at Orono, Orono, Me. 90 pp. Norton, S. A., D. W. Hanson, and J. S. Williams. 1982. Modern and paleolimnological evidence for accelerated leaching and metal accumulation in soils in New England, caused by atmospheric deposition. Water Air Soil Pollut. 18:227-239. Norton, S. A., R. B. Davis, and D. S. Anderson. 1985. The distribution and extent of acid and metal precipitation in Northern New England. Final Report. U.S. Fish and Wildlife Service Grant No. 14-16-0009-75-040 . Norton, V. C., and S. J. Herrmann. 1980. Paleolimnology of fresh-water diatoms (Bacillariophyceae) from a sediment core of a Colorado semidrainage mountain lake. Trans. Am. Microscop. Soc. 99:416-425. Nriagu, J. O. 1983. Arsenic enrichment in lakes near the smelters at Sudbury, Ontario. Geochim. Cosmochim. Acta 47:1523-1526. Nriagu, J. O., and H. H. Harvey. 1978. Isotopic variations as an index of sulfur pollution in lakes around Sudbury, Ontario. Nature 273:223-224.

425 Nriagu, J. O., and H. K. Wong. 1983. Selenium pollution of lakes near the smelters at Sudbury, Ontario. Nature 301:55-57. Nriagu, J. O., A. L. W. Kemp, H. K. T. Wong, and N. Harper. 1979. Sedimentary record of heavy metal pollution in Lake Erie. Geochim. Cosmochim. Acta 43:247-258. Nriagu, J. O., H. K. T. Wong, and R. D. Coker. 1982. Deposition and chemistry of pollutant metals in lakes around the smelters at Sudbury, Ontario. Environ. Sci. Technol. 16:551-560. Nygaard, G. 1956. Ancient and recent flora of diatoms and Chrysophyceae in Lake Gribs¢. Pp. 32-94 of Studies on Humic, Acid Lake Gribs¢, K. Berg and I. C. Peterson, eds. Fol. Limnol. Scand. 8:1-273. Oden, S. 1968. The acidification of air and precipitation and its consequences in the natural environment. Ecology Committee Bull. No. 1, Swedish National Research Council, Stockholm. Oehlert, G. W. 1984. Error bars for paleolimnologically inferred pH histories: a first pass. Technical Report Number 267, Series 2. Department of Statistics, Princeton University, Princeton, N.J. 5 pp. Oldfield, F., R. Thompson, and K. E. Barber. 1978. Changing atmospheric fallout of magnetic particles recorded in recent ombrotrophic peat sections. Science 199:679-680. Oldfield, F., P. G. Appleby, R. S. Cambray, J. D. Eakins, K. E. Barger, R. E. Battarbee, G. R. Pearson, and J. M Williams 1979 210pb 137Cs and 239PU profiles in ombrotrophic peat. Oikos 33:40-45. Oldfield, F., K. Tolonen, and R. Thompson. 1981. History of particulate atmospheric pollution from magnetic measurements in dated Finnish peat profiles. AmbiO 10:185-188. Oliver, B. G., and J. R. M. Kelso. 1983. A role for sediments in retarding the acidification of headwater lakes. Water Air Soil Pollut. 20:379-389. Oliver, B. G., E. M. Thurman, and R. L. Malcolm. 1983. The contribution of humic substances to the acidity of colored natural waters. Geochim. Cosmochim. Acta 47:2031-2035. Olson, E. T. 1983. The distribution and mobility of cesium-137 in Sphagnum bogs of northern North America. M.S. thesis, Department of Physics, University of Maine, Orono, Me. 113 pp.

426 Ottar, B. 1977. International agreement needed to reduce long-range transport of air pollutants in Europe. Ambio 6:262-269. Ouellet, M., and H. G. Jones. 1983. Paleolimnological evidence for the long-range atmospheric transport of acidic pollutants and heavy metals into the Province of Quebec, eastern Canada. Can. J. Earth Sci. 20:23-36. Pakarinen, P. 1981a. Metal content of ombrotrophic Sphagnum mosses in NW Europe. Ann. Bot. Fenn. 18:281-292. Pakarinen, P. 1981b. Nutrient and trace metal content and retention in reindeer lichen carpets of Finnish ombrotrophic bogs. Ann. Bot. Fenn. 18:265-274. Pakarinen, P., and K. Tolonen. 1976. Regional survey of heavy metals in peat mosses (Sphagnum). Ambio 5:38-40. Pakarinen, P., and K. Tolonen. 1977. Distribution of lead in Sphagnum fuscum profiles in Finland. Oikos 28:69-73. Pakarinen, P., K. Tolonen, and J. Soveri. 1980. Distribution of trace metals and sulfur in the surface peat of Finnish raised bogs. Paper presented at the Sixth International Peat Congress. Duluth, Minnesota, Aug. 17-23. Patrick, R. 1943. The diatoms of Linsley Pond, Connecticut. Proc. Acad. Nat. Sci. Philadelphia 95:53-110. Patrick, R. 1954. The diatom flora of Bethany Bog. J. Protozool. 1:34-37. Patrick, R. 1977. Ecology of freshwater diatoms-diatom communities. Pp. 284-332 of The Biology of Diatoms, D. Werner, ed. Botanical Monographs, Volume 13, Oxford, England: Blackwell Scientific Publications. Patrick, R., and C. W. Reimer. 1966. The Diatoms of the United States. Volume 1. Monogr. Acad. Nat. Sci. Philadelphia. 688 pp. Patrick, R., and C. W. Reimer. 1975. The Diatoms of the United States, Volume 2, Part 1. Monogr. Acad. Nat. Sci. Philadelphia. 213 pp. Patrick, R., V. P. Binetti, and S. G. Halterman. 1981. Acid lakes from natural and anthropogenic causes. Science 211:446-448. Patterson, W. A. 1977. The effect of fire on sedimentation in a small lake in Minnesota. Abstract, p. 211 of Program of the Twentieth Congress of the Society of International Limnologists, Copenhagen. Pennington, W. 1981. Records of a lake's life in time: the sediments. Hydrobiologia 79:197-219.

427 Pennington, W. 1984. Long-term natural acidification of upland sites. Pp. 28-46 in Cumbria: Evidence from Post-Glacial Lake Sediments. 52nd Annual Report. Freshwater Biological Association, Ambleside, Cumbria, England. Pennington, W., E. Y. Haworth, A. P. Bonny, and J. P. Lishman. 1972. Lake sediments in northern Scotland. Phil. Trans. R. Soc. London Ser. B 264:191-294. Pennington, W. (Mrs. T. G. Tutin), R. S. Cambray, and E. M. Fisher. 1973. Observations on lake sediments using fallout Cs-137 as a tracer. Nature 242:324-326. Reed, S. 1982. The late-glacial and postglacial diatoms of Heart Lake and Upper Wallface Pond in the Adirondack Mountains, New York. M.S. thesis, San Francisco State University, San Francisco, Calif. Rehakova, Z. 1983. Diatom succession in the post-glacial sediments of the Komorany Lake, northwest Bohemia, Czechoslovakia. Hydrobiologia 103:241-245. Renberg, I. 1976. Paleolimnological investigations in Lake Prastsjon. Early Nor Eland 9:113-159. Renberg, I. 1978. Paleolimnology and verve counts of the annually laminated sediment of Lake Rudetjarn, northern Sweden. Early Norrland 11:63-92. Renberg, I., and T. Hellberg. 1982. The pH history of lakes in southwestern Sweden, as calculated from the subfossil diatom flora of the sediments. Ambio 11:30-33. Renberg, I., and M. Wik. 1984. Dating recent lake sediments by soot particle counting. Verh. Int. Ver. Limnol. 22:712-718. Reuther, R., R. F. Wright, and U. Forstner. 1981. Distribution and chemical forms of heavy metals in sediment cores from two Norwegian lakes affected by acid deposition. Pp. 318-321 of Proceedings of the International Conference on Heavy Metals in the Environment. Edinburgh, United Kingdom: CEP Consultants Ltd. Rochester, H. 1978. Late-glacial and postglacial diatom assemblages of Berry Pond, Massachusetts, in relation to watershed ecosystem development. Ph.D. dissertation, Indiana University, Bloomington, Ind. Rosenqvist, I. Th. 1978. Alternative sources for acidification of river water in Norway. Sci. Total Environ. 10:39-49. Round, F. E. 1957. The late-glacial and post-glacial diatom succession in the Kentmere Valley. New Phytol 56:98-126. .

428 Round, F. E. 1961. The diatoms of a core from Esthwaite Water. New Phytol. 60:43-59. Salomaa, R., and P. Alhonen. 1983. Biostratigraphy of Lake Spitaalijarvi: an ultraoligotrophic small lake in Lauhanvuori, western Finland. Hydrobiologia 103:295-301. Schindler, D. W., R. H. Hesslein, and R. Wagermann. 1980. Effects of acidification on mobilization of heavy metals and radionuclides from the sediments of a freshwater lake. Can. J. Fish. Aquat. Sci. 37:373-377. Schofield, C. L. 1965. Water quality in relation to survival of brook trout, Salvelinus fontinalis (Mitchell). Trans. Am. Fish. Soc. 94:227-235. Schofield, C. L. 1976. Acidification of Adirondack lakes by atmospheric precipitation: extent and magnitude of the problem. Final report, Project F-28-R. New York State Department of Environmental Conservation, Albany, N.Y. Schofield, C. L. lg84. Surface water chemistry in the ILWAS basins. Pp. 6-1--6-32 in The Integrated Lake-Watershed Acidification Study, Vol. 4, R. A. Goldstein and S. A. Gherini, eds. Palo Alto, Calif.: Electric Power Research Institute. Schofield, C. L., and J. N. Galoway. 1977. The utility of paleolimnological analyses of lake sediments for evaluating acid precipitation effects on dilute lakes. Research Project Technical Completion Report. Proj. No. A-071-NY. Office of Water Res. Technol. U.S. Department of the Interior. Sherman, J. W. 1976. Post-pleistocene diatom assemblages in New England lake sediments. Ph.D. dissertation, University of Delaware, Newark, Del. Shiramata, H., R. W. Elias, C. C. Patterson, and M. Koide. 1980. Chronological variations in concentrations and isotopic compositions of anthropogenic atmospheric lead in sediments of a remote subalpine pond. Geocnim. Cosmocnim. Acta 44:149-162. Skei, J., and P. E. Paus. 1979. Surface metal enricument and partitioning of metals in a dated sediment core from a Norwegian fjord. Geochim. Cosmochim. Acta 433:239-246. Smith, R. A. 1872. Air and Rain: The Beginnings of a Chemical Climatology. London: Longmans, Green. Smol, J. P. 1980. Fossil synuracean (Chrysophyceae) scales in lake sediments: a new group of paleoindicators. Can. J. Bot. 58:458-465.

429 Smol, J. P. 1983. Paleophycology of a high arctic lake near Cape Herschel, Ellesmere Island. Can. J. Bot. 61:2195-2204. Smol, J. P. 198Sa. Chrysophycean microfossils as indicators of lakewater pH. In Diatoms and Lake Acidity, J. P. Smol, R. W. Battarbee, R. B. Davis, and J. Merilainen, eds. The Hague, The Netherlands: W. Junk. Smol, J. P. 1985b. Chrysophycean microfossils in paleolimnological studies. Palaeogeogr. Palaeoclimatol. Palaeoecol. (In press.) Smol, J. P., and M. D. Dickman. 1981. The recent histories of three Canadian Shield lakes: a paleolimnological experiment. Arch. Hydrobiol. 93:83-108. Smol, J. P., D. F. Charles, and D. R. Whitehead. 1984a. Mallomonadacean (Chrysophyceae) assemblages and their relationships with limnological characteristics in 38 Adirondack (New York) lakes. Can. J. Bot. 62:911-923. Smol, J. P., D. F. Charles, and D. R. Whitehead. 1984b. Mallomonadacean microfossils provide evidence of recent lake acidification. Nature 307:628-630. Smol, J. P., R. W. Battarbee, R. B. Davis, and J. Merilainen, eds. 1985. Diatoms and Lake Acidity. The Hague, The Netherlands: W. Junk. (In press.) Somers, K. M., and H. H. Harvey. 1984. Alteration of fish communities in lakes stressed by acid deposition and heavy metals near Wawa, Ontario. Can. J. Fish. Aguat. Sci. 41:20-29. Sreenivasa, M. R., and H. C. Duthie. 1973. The postglacial diatom history of Sunfish Lake, southwestern Ontario. Can. J. Bot. 51:1599-1609. Steinnes, E. 1977. Atmospheric deposition of trace elements in Norway studied by means of moss analysis. Kjeller, Norway: Institute for Atomenergi. 13 pp. and Appendix. Stockner, J. G., and W. W. Benson. 1967. The succession of diatom assemblages in the recent sediments of Lake Washington. Limnol. Oceanogr. 12:513-533. Sweets, P. R. 1983. Differential deposition of diatom frustules in Jellison Hill Pond, Maine. M.S. thesis, University of Maine, Orono, Me. 156 pp. Tan, Y. L., and M. Heit. 1981. Biogenic and abiogenic polynuclear aromatic hydrocarbons in sediments from two remote Adirondack Lakes. Geochim. Cosmochim. Acta 45:2267-2279

430 Taylor, D. R., M. A. Tompkins, S. E. Kirton, T. Mauney, and D. F. S. Natusch. 1982. Analysis of fly ash produced from combustion of refuse-derived fuel and coal mixtures. Environ. Sci. Technol. 16:148-154. Tolonen, K. 1967. Uber die Entwicklung der Moore in finnischen Nordbarelien. Ann. Bot. Fenn. 4:219-416. Tolonen, K. 1972. On the palaeo-ecology of the Hamptjarn Basin II. Bio- and chemostratigraphy. Early Norrland 1:53-77. Tolonen, K. 1980. Pollen, algal remains, and macrosub- fossils from Lake Galltrask, S. Finland. Ann. Bot. Fenn. 17:394-405. Tolonen, K., and T. Jaakkola. 1983. History of lake acidification and air pollution studied in sediments in South Finland. Ann. Bot. Fenn. 20:57-78. Tolonen, K., and J. Merilainen. 1983. Sedimentary chemistry of a small polluted lake, Galltrask, S. Finland. Hydrobiologia 103:309-318. Tolonen, K., R. B. Davis, and L. S. Widoff. 1984. Peat accumulation rates in selected Maine peatlands. Maine Geological Survey, Department of Conservation. 109 pp. Tolonen, K., M. Liukkonen, R. Harjula, and A. Patila. 1985. Acidification of small lakes in Finland studied by means of sedimentary diatom and chrysophyceae remains. In Diatoms and Lake Acidity, J. P. Smol, R. W. Battarbee, R. B. Davis, and J. Merilainen, eds. The Hague, The Netherlands: W. Junk. Tolonen, K., R. B. Davis, and L. Widoff. Peat accumlation rates in selected Maine peatlands. Maine Geol. Surv. Bull. (In press.) Tynni, R. 1972. The development of Lovojarvi on the basis of its diatoms. Aqua Fenn. 74-82. van Dam, H., and H. Kooyman-van Blokland. 1978. Man-made changes in some Dutch moorland pools, as reflected by historical and recent data about diatoms and macrophytes. Int. Rev. Gesam. Hydrobiol. 63:587-607 van Dam, H., G. Suurmond, and C. J. F. ter Braak. 1981. Impact of acidification on diatoms and chemistry of Dutch moorland pools. Hydrobiologia 83:425-459. Vitt, D. H., and S. Bayley. 1984. The vegetation and water cnemistry of four oligotrophic basin mires in northwestern Ontario. Can. J. Bot. 62:1485-1500. Wakeham, S. G., C. Schaffner, and W. Giger. 1980. Polycyclic aromatic hydrocarbons in recent lake sediments. I. Compounds having anthropogenic or~gins. Geochim. Cosmochim. Acta 44:403-413. .

431 Walker, R. 1978. Diatom and pollen studies of a sediment profile from Melynllyn, a mountain tarn in Snowdownia, North Wales. New Phytol. 81:791-804. Watanabe, T., and I. Yasuda. 1982. Diatom assemblages in lacustrine sediments of Lake Shibu-ike, L. Misume-ike, L. Naga-ike, L. Kido-ike in Shiga Highland and a new biotic index based on the diatom assemblage for the acidity of lake water. Jpn. J. Limnol. 43:237-245. Whitehead, D. R., D. F. Charles, S. E. Reed, S. T. Jackson, and M. C. Sheehan. In press. Late-glacial and Holocene acidity changes in Adirondack (N.Y.) lakes. ~ In Diatoms and Lake Acidity, J. P. Smol, R. W. Battarbee, R. B. Davis, and J. Merilainen, eds. Dordrecht, The Netherlands: W. Junk. Wright, H. E., Jr. 1981. The role of fire in land/water interactions. In Fire Regimes and Ecosystem Properties, Pp. 421-444 of Proceedings of a conference held December 11-15, 1978, in Honolulu, Hawaii. General Technical Report W0-26, U.S. Department of Agriculture. Wright, R. F. 1983. Predicting acidification of North American lakes. Report 4/1983, Norwegian Institute for Water Research, Oslo. 165 pp. Wright, R. F. 1984. Changes in the chemistry of Lake Hovvatn, Norway following liming and reacidification. Report 6/1984, Norwegian Institute for Water Research, Oslo. 68 pp. Yan, N. 1983. Effects of changes in pH on transparency and thermal regimes of Lohi lake, near Sudbury, Ontario. Can. J. Fish. Aquat. Sci. 40:621-626.

APPENDIXES